CA1056060A - Method and apparatus for enhancing data - Google Patents
Method and apparatus for enhancing dataInfo
- Publication number
- CA1056060A CA1056060A CA243,103A CA243103A CA1056060A CA 1056060 A CA1056060 A CA 1056060A CA 243103 A CA243103 A CA 243103A CA 1056060 A CA1056060 A CA 1056060A
- Authority
- CA
- Canada
- Prior art keywords
- data
- elements
- digitized
- enhancing
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/30—Transforming light or analogous information into electric information
- H04N5/32—Transforming X-rays
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/40—Picture signal circuits
- H04N1/407—Control or modification of tonal gradation or of extreme levels, e.g. background level
- H04N1/4072—Control or modification of tonal gradation or of extreme levels, e.g. background level dependent on the contents of the original
Abstract
ABSTRACT OF THE DISCLOSURE
A technique and apparatus for enhancing the data upon a film transparency is presented. The density characteristics of the object film are digitized and shifted to exponentially increase the magnitude and significance of the digitized data. This amplified data is then utilized for producing a second film transparency having increased contrast in accordance with the ampli-fication of the original data. There is also provided means for offsetting the digitized data with respect to a fixed reference such that features of interest may be separated from their background before the final recording of the enhanced film.
A technique and apparatus for enhancing the data upon a film transparency is presented. The density characteristics of the object film are digitized and shifted to exponentially increase the magnitude and significance of the digitized data. This amplified data is then utilized for producing a second film transparency having increased contrast in accordance with the ampli-fication of the original data. There is also provided means for offsetting the digitized data with respect to a fixed reference such that features of interest may be separated from their background before the final recording of the enhanced film.
Description
~o~ o METHOD AND APPARATUS FOR ENHANCING DATA
~ACK~ROUND OF 1~ INVRNTION
The utilization of x-rays and other photographic medium for obtaining data relative to the structural characteristics of an object has been known for many years.
Images are often encountered wherein the features of interest may be separated from their surroundings by only a small difference in brightness, thus making the exam-ination or even the detection of these features a difficult proposition for a human inspecting the same. The images of interest may be in an electronic form such as the video information being fed to a closed circuit television -~ monitor, or in photographic form such as in x-ray or nuclear radiograph of a part being inspected for cracks `~
and other flaws. Of course, medical radiographs and the interpretation thereof are also sub~ect to such problems.
An additional difficulty in the case of radiographs stems from the somewhat limited ability of the human eye to observe minute density differences in those portions of the photographic transparencies with near zero density , (almost transparent) and in portions with high density ,i (almost no light transmittance).
3 In order to make maximum use of the photographic I images discussed hereinabove, a device and technique must ¦ 25 be implemented ~or enhancing the contrast of an image at all levels of density (or brightness) without exceeding the dynamic range of the display medium (cathode ray tube monitor, or film) and without exceeding the dynamic range of the human utilizing the same.
: _.
: ~ ~05~
Consequently, it is an object of the instant invention to present a method and apparatus for enhancing data whereby data recorded on a first media in accordance with a first transfer function may again be recorded on a second medium in accordance with a second transfer function, the second transfer function being an enhancement of the first. ' Yet another object of the invention is to present a method and apparatus for enhancing data whereby imaged data may be digitized, the digitized values shifted, and re-imaged so as to present an enhanced image structure.
Still another object of the invention is to present a method and apparatus for enhancing data whereby image data may be digitized, and the digitized data be shifted 15 and offset an incremental amount so as to provide for maximum enhancement between background structure and subtle features.
' Still another object of the invention is to present ~ a method and apparatus for enhancing data whereby the positioning and size of a flaw within an object may be determined.
, Still a further object of the invention is to '. present a method and apparatus for enhancing data which is , inexpensive, reliable, and easily developed from the state of the art.
~" Thes'e'objects and other objects which will become `~ apparent as the'detailed description proceeds are achieved by an apparatus for enhancing the data contained on a film
~ACK~ROUND OF 1~ INVRNTION
The utilization of x-rays and other photographic medium for obtaining data relative to the structural characteristics of an object has been known for many years.
Images are often encountered wherein the features of interest may be separated from their surroundings by only a small difference in brightness, thus making the exam-ination or even the detection of these features a difficult proposition for a human inspecting the same. The images of interest may be in an electronic form such as the video information being fed to a closed circuit television -~ monitor, or in photographic form such as in x-ray or nuclear radiograph of a part being inspected for cracks `~
and other flaws. Of course, medical radiographs and the interpretation thereof are also sub~ect to such problems.
An additional difficulty in the case of radiographs stems from the somewhat limited ability of the human eye to observe minute density differences in those portions of the photographic transparencies with near zero density , (almost transparent) and in portions with high density ,i (almost no light transmittance).
3 In order to make maximum use of the photographic I images discussed hereinabove, a device and technique must ¦ 25 be implemented ~or enhancing the contrast of an image at all levels of density (or brightness) without exceeding the dynamic range of the display medium (cathode ray tube monitor, or film) and without exceeding the dynamic range of the human utilizing the same.
: _.
: ~ ~05~
Consequently, it is an object of the instant invention to present a method and apparatus for enhancing data whereby data recorded on a first media in accordance with a first transfer function may again be recorded on a second medium in accordance with a second transfer function, the second transfer function being an enhancement of the first. ' Yet another object of the invention is to present a method and apparatus for enhancing data whereby imaged data may be digitized, the digitized values shifted, and re-imaged so as to present an enhanced image structure.
Still another object of the invention is to present a method and apparatus for enhancing data whereby image data may be digitized, and the digitized data be shifted 15 and offset an incremental amount so as to provide for maximum enhancement between background structure and subtle features.
' Still another object of the invention is to present ~ a method and apparatus for enhancing data whereby the positioning and size of a flaw within an object may be determined.
, Still a further object of the invention is to '. present a method and apparatus for enhancing data which is , inexpensive, reliable, and easily developed from the state of the art.
~" Thes'e'objects and other objects which will become `~ apparent as the'detailed description proceeds are achieved by an apparatus for enhancing the data contained on a film
2~
'~
, . " . . - - .- . . -1(~5~ 0 transparency, characterized by: first means for scanning ' said film transparency in discrete elements and digitizing the data of each of said elements; a shifting circuit connected to said first means for shifting the digitized data o~ each and every element a predetermined number of positions toward the most significant bit of said data;
second means connected to said shifting circuit for convert-ing said amplified digital data into analog signals; and third means connected to said second means for receiving said analog signals and recording the same on an output medium.
For a complete understanding of the structure and techniques of the invention, reference should be had to the following detailed description and accompanying drawings wherein:
Fig. 1 is a graphical illustration of the standard transfer function associated with image production;
Fig. 2 is a gràphical illustration of a logarithmic -' transfer function, providing a first step of enhancement;
Fig. 3 is a graphical illustration of the transfer function obtained with a large gain in the system;
Fig. 4 is a graphical illustration of a transfer ' function resulting in an image enhancement of 16x;
', Fig. 5 is a graphical illustration of the enhanced , 25 transfer function of Fig. 4 having a 25% offset;
,~ Fig. 6 is a block diagram schematic of the apparatus of the invention~; and Fig. 7 is a block diagram schematic of the gradient removal circuitry of the invention as the same is incorporated ~3~
. .
lQSjL~O~O
in the structure of Fig. 6.
A normal photo transparency, such as a radiograph or x-ray, contains upon a two dimensional reference, a plurality of areas of varying light transmittance charac-teristics. These variations are indicative of the parti-cular characteristics of the imaged object but often times the most important data acquired by means of the x-ray or ,. 10 -3a-, .
lO5SO~iO
radiograph is hidden or undiscernable within the back-ground of the structure of the object. However, the information or critical data is indeed present on the -~
image-bearing transparency but is only unintelligible in that state. Consequently, it becomes necessary to re-produce the image of the transparency in a unique and unusual way in order to enhance those areas which bear the critical data.
If the original image were to be reproduced with no enhancement whatsoever, the transfer function (bright-ness out vs. brightness in, or density out vs. density in) would be a straight line function as shown in Fig. 1.
The reproduced image would be an exact replica of the original image due to the straight line transfer function.
The first step in enhancing the original image in the reproduction process is to use a logarithmic transfer ; function as shown in Fig. 2 where the input is in terms of density of the original image and the output is in terms of exposure on film stock to generate a new transparency.
; 20 A~ is well known in the art, density is a log function of transmission and thus, even though Fig. 2 plots as a straight line, the same is a logarithmic function since density i8 a logarithmic function of the transmittance or brightness of the original image. While the graph of Fig. 2 would indicate that a reversal would be present when developing the reproduction, with original blacks coming out white and vice versa, this is of no conse-quence since the output media may contain yet another reversal if necessary by, for example, making the final . .
., 1055~060 exposure on normal film stock.
As is well known in the art, the process of expos-ing an original radiograph inherently contains an expon-ential transfer function. Consequently, by using the logarithmic function of Fig. 2 in the enhancement process, the transmittance of the enhanced image becomes linearly related to the thickness of the original part and leads to the conclusion that the enhanced image may be calibrated to determine the actual thickness at any point on the original part.
By using the transfer function as related in Fig. 2, no vast improvement or increase in contrast is achieved.
Fig. 3 presents the transfer function obtained if a large gain is placed in the reproduction system. The contrast here is greatly enhanced but only over a small range of the input density and the output media quickly saturates.
In other woras, those areas of the original image of den8ities of 0 - lD are greatly enhanced in the reproduc-tion but those areas of density greater than lD become unintelligible in the reproduction.
In Fig. 4 there i9 presented an enhancement transfer function which presents a great increase in contrast over the full input dynamic range. The enhancement utilizing thl8 density tran8fer function i~ 16 times that of the original image enhancement. This transfer function appear~
as an idealized saw-tooth wave having a positive slope variable in dependence upon the number of "teethn 4elected to cover the full input density range. The return or negative slope of the transfer function is always minus .;~ .
, _5-, lVS~O~O
infinity. By utilizing the transfer function of Fig. 4, contrast may be greatly enhanced while all geographic areas of the original image will appear with enhanced contrast on the output image since any input density value has a valid output value. This is true even if the original image had an extremely large dynamic range.
A further improvement may be made upon the transfer function related in Fig. 4 by shifting the 0 point of the saw-teeth laterally along the input density range allowing the discontinuities of the saw-teeth to be positioned at any selected input density as shown in Fig. 5. It should be readily apparent to those skilled in the art that the greatest feature enhancement is then found when the den~ity ; of the feature being sought and the density of the surround-ing area are split by one of the saw-~eeth vertical dlscontinuities. In other word~, at almost adjacent points along the input range, an almost total separation on the output range will be experienced. This gives a near infinite gamma or contrast and the output image then shows the feature as white on a black background or vice versa.
Of course, the lateral shifting of the transfer function i8 only an increment of the saw-tooth period since greater ~hifts would be redundant.
.; It ~hould be readily apparent then that by utilizing ; 25 an original x-ray, radiograph,or other transparency as an ` input image, a greatly enhanced output image may be achieved by means of the utilization of the transfer function of Figs. 4 and 5.
Referring now to Fig. 6, the structures necessary ~, .
. .
.~, ~ -6-'I' " :
.' . ' ' 105~0~0 for generating and implementing the transfer functions discussed hereinabove can be seen. The original image is contained upon the film 12 which may comprise an x-ray, radiograph~or other suitable transparency. ~he film 12 is appropriately scanned by means of the input scanner 14 which may comprise a drum scanner, microdensitometer, vidicon tube, image dissector tube, or the like. Suitable devices which have been utilized include the Model 57 image digitizer by Dicomed Inc. and the scanning micro-densitometer, Model 650, by Technical Operations, Incor-porated of Burlington, Massachusetts. In either case, these devices are well known and understood by those skilled in the art and are used for purposes of sequenti-ally scanning the input image on an element by element basis, The~e picture elements, commonly called pixels, ; are each individually and sequentially scanned by scanner 14, The size of the pixels and the spacings thereof are of course programmable and, indeed, the pixels may be spaced apart or may overlap in accordance with the dictates of the operator. If the input scanner 14 is operative for producing an output corresponding to light transmi~sion intensity through the film 12, then it is necessary to pass the output thereof through the logarithmic amplifier 16 such that the output of the amplifier 16 is indicative of the density of the various film pixels. As mentioned hereinabove, and as i8 well known to those skilled in the art, density is a logarithmic function of intensity and hence the log amplifier 16 provides an important function for those devices producing outputs related to intensity.
.
' :
~ .
:~ . - . - . .. :
~ ~ 5 ~O ~ O
It should of course be understood that many devices present outputs corresponding to density and hence the log amplifier 16 would not be required. In any event, there is applied to an analog to digital converter 18 analog signals indicative of the density of the related pixels being scanned.
There is presented at the output of the analog to digital converter 18 digital signals relating the density of the associated pixels; thus, the data on the film 12 is digitized. The output of the converter 18 is passed to a full adder 20 which also receives an input from the offset selector 32. For purposes of the present discussion, it will be assumed that there is no input supplied by the offset selector 32 and hence the output of the full adder 20 is the same as the digitized input thereto rec-eived from the converter 18.
me shifter 22 functions to achieve the generation `~ of the saw-tooth portion of the transfer function discussed hereinabove. This element shifts the outputs of the full adder 22 toward the most significant bit before applying them to the input of the digital to analog converter 24. In other words, the shifter 22 is operative to delete a num-ber of the mo~t significant bits received from the full adder 20 and shift all of the remaining bits the number of positions deleted thus amplifying the significance of the remaininq bits. Of course, the shifter 22 could comprise nothing more than a switch operative for selec-~ tively connecting the outputs of the full adder 20 to `~ the input~ of the converter 2~. It should be readily .' . .
:,'' ~ ' ....
' -8-~. . ... ~ . .. - . - . ..
l~S~IV~O
appreciated that a shift of 1 bit results in a ~oubling of the significance of the remaining bits whi]e a shift of two bits quadruples the significance of the remaining bits and so on. In other words, if the shifter 22 shifts the data n bits then the significance of the remaining bits is increased by 2n. It is important to note that the shift from the outputs of the full ad~er ~0 to the inputs of the digital to analog converter 24 is toward the - most significant bit.
There is thus presented at the output of the con-verter 24 an amplified analog signal enhanced by means of the shifter 22 and bearing the data present on the film 12.
This analog signal is then passed to a standard film recorder 26 whereby the development of an enhanced film 28 may be achieved by the normal process. Of course, the signal could be passed to an appropriate monitor or C~T
30 in a normal fashion.
As discu~sed hereinabove with respect to Fig. 5, the ~aw-tooth transfer function may be laterally offset with respect to the input data values presented to the shifter 22 by mean~ of the addition of an offset increment.
By adding a fixed value selected by means of the offset selector 32 to the output of the analog to digital converter 18 within the full adder 20, thi~ offset may be achieved.
The purpose of this offset again is to force the features ~eing sought to fall on nearly afljacent opposite sides of the negative going slope of the saw-tooth curve such that the contrast between the feature and the background will be optimized. Various increments of offset may be selected .
.
.
1~35~V~O
and various passes made on each of the image films 1~ to swing through the full range covered by a period of the transfer function to guarantee that optimum contrast has been achieved for each object film or each geographical area of a given object film.
In general than it can be seen that the technique and apparatus of the invention is operative for initially digitizing the density-related signal achieved from scanning an object film, offsettin~ that digiti~ed value, shifting the values towards the most significant bit to achieve the desired enhancement, and reconverting the signals to an analog signal, if need be, for reproduction in a common film recorder of an enhanced film.
A further attribute of the technique and utilization of the structure of the instant invention as described above resides in the ability to produce a quasi-three dimensional view of an object from a two dimensional transparency of that object. By referrin~ to the transfer function of Fig. 4, it can be appreciated that the sawtooth transfer function shown therein actually comprises sixteen individual ramp transfer functions separated by discontinuities.
In other words, each period of the transfer function could be treated as a separate and distinct transfer function . ~ .
`- and allowed to enhance an object film in accordance therewith.
t 25 If each such period of the transfer function of Fig. ~
we~e so treated, an enhancement of an object film could be achieved on a composite of sixteen enhanced output media.
Each of the output media operates on a different range of input densities enhancing the same to the output densities ~, .
.:
,~'`~ .
,i, , , ~ .
~Q~ O
correlating to the function of Fig. 4. As stated here-inbefore, the transmittance of the enhanced image is linearly - related to the thickness of the original part and hence, by stacking the plurality of output media on top of each other in order of the period of the transfer function which they represent, a quasi three dimensional view of the object film may be achieved.
Even utilizing the techniques and structure of the invention presented hereinabove, the distinguishing of flaws, cracks or the like within the structural background of a work piece may often times be a difficult proposition.
It has been found that the contours of most flaws, and particularly cracks, change more rapidly than do the natural contours of a work piece. Consequently, by normalizing the transparency and each pixel thereof with respect to the average of all neighboring pixels it has been found that flaws may be readily discernable. A circuit for achieving such averaging or normalization is shown in Fig. 6 as the element 40 which may be interconnected between the A/D converter 18 and full adder 20. A block diagram of the structure of this element 40 is shown in Fig 7.
Referring now to Fig. 7, it can he seen that the digltizea pixel information from the A/~ converter lR
is supplied to a shift register 42; such s~ift register being capable of containing information rel~tive to 32 such pixels. Connected to and receiving the output of the shift register 42 is a similar shift register 44, again being a 32 s~age register. An accumulator 46 is . --11--.- -- : . - -~.~: -. - . . . :
~QS`~060 provided to receive and maintain an update~ a~gre~ate of the digitized pixel densities as the ~e are received from the A/D converter 18. As the microdensitometer is utilized in analyzing th~ ima~e, it is ~rocessed on S a pixel by pixel basis, such that the value within the accumulator 4h is updated by adding to the present value thereof the value of the new pixel within the ad~er 4R.
This updated value is then reduced by the value of the pixel 64 pixels away within the subtract circuit 5n. This value is then storea in the accumulator ~6. ~asically, the accumulator 46 contains an a~re~ate of the digitized values of 64 pixels, the value being updated by replacing the value of the pixel being scanned at any particular time for the value of the pixel scanned 6~ pixels earlier.
The output of the accumulator 46 is passed to the divider 52 to give an average of the density of the 6~
pixels stored therein. The divi~er 52 may, of course, be of any suitable nature as is well understoo~ ~y those skilled in ~he art. The pixel average output of the divider 52 is passed to the su~tract circuit 54 which al~o receives the output of the shift re~ister 42. It should be readily . ~ . .
~` apparent that this output is the di~iti~ed value of the ~` center plxel of that group o~ pixels representea hy the value stored within the accumulator 46. The OUtpllt of ~, 25 the subtract circuit 54 is thus the devi~tion of the density i of a pixel from the value of surroundin~ pixels on either side thereof. To a de~ree, the output of the element 54 ~; is thus a normalized den~ity output. This OlltpUt may be applied directly to the input of the full adder ~0 as shown ., `':
~ ~ .
"~ , .. .
1.0~ 0 in Fi~. 7. However, it is often times undesirAhle to have a negative input to the adder 2~ and hence the output of subtract circuit 54 is applied to an adder 5~ wherein the same is altered by a bias level. The bias applie~ to the adder 56 is most generally selected as the center of the operating range of pixel deviation such that positive and negative excursions about the bias level will be substantial-~y equal. The bias level virtually guarantees that the output of the adder 5~ will cause all inputs to the adder 20 to be positive.
It should be readily apparent from an understanding of the function of the circuitry of Fia. 7, that the hi~h frequency changes of density associatecl with flaws, cracks and the like will result in highly distinquishable de-viations above the average value of nei~hboring pixels as the same are contained in the accumulator ~. Thus, the normalized output resulting from the structure of Fig. 7 and combined with the offset techni~ue as shown in Fig. 5 will readily distinguish sharply var~ina density deviations above the background density.
Thus it can be seen that the objects of the invention have been satisfied by the techniques and structure presented hereinabove. ~Ihile in accordance with the patent statutes only the ~est mo~e and preferred embodiment of the invention has been presented and des-cribed in detail, it is to be understood that the invention is not limited thereto or thereby. Conse~uently, for an appreciation of the scope and breadth of the invention reference should be had to the followin~ claims.
'
'~
, . " . . - - .- . . -1(~5~ 0 transparency, characterized by: first means for scanning ' said film transparency in discrete elements and digitizing the data of each of said elements; a shifting circuit connected to said first means for shifting the digitized data o~ each and every element a predetermined number of positions toward the most significant bit of said data;
second means connected to said shifting circuit for convert-ing said amplified digital data into analog signals; and third means connected to said second means for receiving said analog signals and recording the same on an output medium.
For a complete understanding of the structure and techniques of the invention, reference should be had to the following detailed description and accompanying drawings wherein:
Fig. 1 is a graphical illustration of the standard transfer function associated with image production;
Fig. 2 is a gràphical illustration of a logarithmic -' transfer function, providing a first step of enhancement;
Fig. 3 is a graphical illustration of the transfer function obtained with a large gain in the system;
Fig. 4 is a graphical illustration of a transfer ' function resulting in an image enhancement of 16x;
', Fig. 5 is a graphical illustration of the enhanced , 25 transfer function of Fig. 4 having a 25% offset;
,~ Fig. 6 is a block diagram schematic of the apparatus of the invention~; and Fig. 7 is a block diagram schematic of the gradient removal circuitry of the invention as the same is incorporated ~3~
. .
lQSjL~O~O
in the structure of Fig. 6.
A normal photo transparency, such as a radiograph or x-ray, contains upon a two dimensional reference, a plurality of areas of varying light transmittance charac-teristics. These variations are indicative of the parti-cular characteristics of the imaged object but often times the most important data acquired by means of the x-ray or ,. 10 -3a-, .
lO5SO~iO
radiograph is hidden or undiscernable within the back-ground of the structure of the object. However, the information or critical data is indeed present on the -~
image-bearing transparency but is only unintelligible in that state. Consequently, it becomes necessary to re-produce the image of the transparency in a unique and unusual way in order to enhance those areas which bear the critical data.
If the original image were to be reproduced with no enhancement whatsoever, the transfer function (bright-ness out vs. brightness in, or density out vs. density in) would be a straight line function as shown in Fig. 1.
The reproduced image would be an exact replica of the original image due to the straight line transfer function.
The first step in enhancing the original image in the reproduction process is to use a logarithmic transfer ; function as shown in Fig. 2 where the input is in terms of density of the original image and the output is in terms of exposure on film stock to generate a new transparency.
; 20 A~ is well known in the art, density is a log function of transmission and thus, even though Fig. 2 plots as a straight line, the same is a logarithmic function since density i8 a logarithmic function of the transmittance or brightness of the original image. While the graph of Fig. 2 would indicate that a reversal would be present when developing the reproduction, with original blacks coming out white and vice versa, this is of no conse-quence since the output media may contain yet another reversal if necessary by, for example, making the final . .
., 1055~060 exposure on normal film stock.
As is well known in the art, the process of expos-ing an original radiograph inherently contains an expon-ential transfer function. Consequently, by using the logarithmic function of Fig. 2 in the enhancement process, the transmittance of the enhanced image becomes linearly related to the thickness of the original part and leads to the conclusion that the enhanced image may be calibrated to determine the actual thickness at any point on the original part.
By using the transfer function as related in Fig. 2, no vast improvement or increase in contrast is achieved.
Fig. 3 presents the transfer function obtained if a large gain is placed in the reproduction system. The contrast here is greatly enhanced but only over a small range of the input density and the output media quickly saturates.
In other woras, those areas of the original image of den8ities of 0 - lD are greatly enhanced in the reproduc-tion but those areas of density greater than lD become unintelligible in the reproduction.
In Fig. 4 there i9 presented an enhancement transfer function which presents a great increase in contrast over the full input dynamic range. The enhancement utilizing thl8 density tran8fer function i~ 16 times that of the original image enhancement. This transfer function appear~
as an idealized saw-tooth wave having a positive slope variable in dependence upon the number of "teethn 4elected to cover the full input density range. The return or negative slope of the transfer function is always minus .;~ .
, _5-, lVS~O~O
infinity. By utilizing the transfer function of Fig. 4, contrast may be greatly enhanced while all geographic areas of the original image will appear with enhanced contrast on the output image since any input density value has a valid output value. This is true even if the original image had an extremely large dynamic range.
A further improvement may be made upon the transfer function related in Fig. 4 by shifting the 0 point of the saw-teeth laterally along the input density range allowing the discontinuities of the saw-teeth to be positioned at any selected input density as shown in Fig. 5. It should be readily apparent to those skilled in the art that the greatest feature enhancement is then found when the den~ity ; of the feature being sought and the density of the surround-ing area are split by one of the saw-~eeth vertical dlscontinuities. In other word~, at almost adjacent points along the input range, an almost total separation on the output range will be experienced. This gives a near infinite gamma or contrast and the output image then shows the feature as white on a black background or vice versa.
Of course, the lateral shifting of the transfer function i8 only an increment of the saw-tooth period since greater ~hifts would be redundant.
.; It ~hould be readily apparent then that by utilizing ; 25 an original x-ray, radiograph,or other transparency as an ` input image, a greatly enhanced output image may be achieved by means of the utilization of the transfer function of Figs. 4 and 5.
Referring now to Fig. 6, the structures necessary ~, .
. .
.~, ~ -6-'I' " :
.' . ' ' 105~0~0 for generating and implementing the transfer functions discussed hereinabove can be seen. The original image is contained upon the film 12 which may comprise an x-ray, radiograph~or other suitable transparency. ~he film 12 is appropriately scanned by means of the input scanner 14 which may comprise a drum scanner, microdensitometer, vidicon tube, image dissector tube, or the like. Suitable devices which have been utilized include the Model 57 image digitizer by Dicomed Inc. and the scanning micro-densitometer, Model 650, by Technical Operations, Incor-porated of Burlington, Massachusetts. In either case, these devices are well known and understood by those skilled in the art and are used for purposes of sequenti-ally scanning the input image on an element by element basis, The~e picture elements, commonly called pixels, ; are each individually and sequentially scanned by scanner 14, The size of the pixels and the spacings thereof are of course programmable and, indeed, the pixels may be spaced apart or may overlap in accordance with the dictates of the operator. If the input scanner 14 is operative for producing an output corresponding to light transmi~sion intensity through the film 12, then it is necessary to pass the output thereof through the logarithmic amplifier 16 such that the output of the amplifier 16 is indicative of the density of the various film pixels. As mentioned hereinabove, and as i8 well known to those skilled in the art, density is a logarithmic function of intensity and hence the log amplifier 16 provides an important function for those devices producing outputs related to intensity.
.
' :
~ .
:~ . - . - . .. :
~ ~ 5 ~O ~ O
It should of course be understood that many devices present outputs corresponding to density and hence the log amplifier 16 would not be required. In any event, there is applied to an analog to digital converter 18 analog signals indicative of the density of the related pixels being scanned.
There is presented at the output of the analog to digital converter 18 digital signals relating the density of the associated pixels; thus, the data on the film 12 is digitized. The output of the converter 18 is passed to a full adder 20 which also receives an input from the offset selector 32. For purposes of the present discussion, it will be assumed that there is no input supplied by the offset selector 32 and hence the output of the full adder 20 is the same as the digitized input thereto rec-eived from the converter 18.
me shifter 22 functions to achieve the generation `~ of the saw-tooth portion of the transfer function discussed hereinabove. This element shifts the outputs of the full adder 22 toward the most significant bit before applying them to the input of the digital to analog converter 24. In other words, the shifter 22 is operative to delete a num-ber of the mo~t significant bits received from the full adder 20 and shift all of the remaining bits the number of positions deleted thus amplifying the significance of the remaininq bits. Of course, the shifter 22 could comprise nothing more than a switch operative for selec-~ tively connecting the outputs of the full adder 20 to `~ the input~ of the converter 2~. It should be readily .' . .
:,'' ~ ' ....
' -8-~. . ... ~ . .. - . - . ..
l~S~IV~O
appreciated that a shift of 1 bit results in a ~oubling of the significance of the remaining bits whi]e a shift of two bits quadruples the significance of the remaining bits and so on. In other words, if the shifter 22 shifts the data n bits then the significance of the remaining bits is increased by 2n. It is important to note that the shift from the outputs of the full ad~er ~0 to the inputs of the digital to analog converter 24 is toward the - most significant bit.
There is thus presented at the output of the con-verter 24 an amplified analog signal enhanced by means of the shifter 22 and bearing the data present on the film 12.
This analog signal is then passed to a standard film recorder 26 whereby the development of an enhanced film 28 may be achieved by the normal process. Of course, the signal could be passed to an appropriate monitor or C~T
30 in a normal fashion.
As discu~sed hereinabove with respect to Fig. 5, the ~aw-tooth transfer function may be laterally offset with respect to the input data values presented to the shifter 22 by mean~ of the addition of an offset increment.
By adding a fixed value selected by means of the offset selector 32 to the output of the analog to digital converter 18 within the full adder 20, thi~ offset may be achieved.
The purpose of this offset again is to force the features ~eing sought to fall on nearly afljacent opposite sides of the negative going slope of the saw-tooth curve such that the contrast between the feature and the background will be optimized. Various increments of offset may be selected .
.
.
1~35~V~O
and various passes made on each of the image films 1~ to swing through the full range covered by a period of the transfer function to guarantee that optimum contrast has been achieved for each object film or each geographical area of a given object film.
In general than it can be seen that the technique and apparatus of the invention is operative for initially digitizing the density-related signal achieved from scanning an object film, offsettin~ that digiti~ed value, shifting the values towards the most significant bit to achieve the desired enhancement, and reconverting the signals to an analog signal, if need be, for reproduction in a common film recorder of an enhanced film.
A further attribute of the technique and utilization of the structure of the instant invention as described above resides in the ability to produce a quasi-three dimensional view of an object from a two dimensional transparency of that object. By referrin~ to the transfer function of Fig. 4, it can be appreciated that the sawtooth transfer function shown therein actually comprises sixteen individual ramp transfer functions separated by discontinuities.
In other words, each period of the transfer function could be treated as a separate and distinct transfer function . ~ .
`- and allowed to enhance an object film in accordance therewith.
t 25 If each such period of the transfer function of Fig. ~
we~e so treated, an enhancement of an object film could be achieved on a composite of sixteen enhanced output media.
Each of the output media operates on a different range of input densities enhancing the same to the output densities ~, .
.:
,~'`~ .
,i, , , ~ .
~Q~ O
correlating to the function of Fig. 4. As stated here-inbefore, the transmittance of the enhanced image is linearly - related to the thickness of the original part and hence, by stacking the plurality of output media on top of each other in order of the period of the transfer function which they represent, a quasi three dimensional view of the object film may be achieved.
Even utilizing the techniques and structure of the invention presented hereinabove, the distinguishing of flaws, cracks or the like within the structural background of a work piece may often times be a difficult proposition.
It has been found that the contours of most flaws, and particularly cracks, change more rapidly than do the natural contours of a work piece. Consequently, by normalizing the transparency and each pixel thereof with respect to the average of all neighboring pixels it has been found that flaws may be readily discernable. A circuit for achieving such averaging or normalization is shown in Fig. 6 as the element 40 which may be interconnected between the A/D converter 18 and full adder 20. A block diagram of the structure of this element 40 is shown in Fig 7.
Referring now to Fig. 7, it can he seen that the digltizea pixel information from the A/~ converter lR
is supplied to a shift register 42; such s~ift register being capable of containing information rel~tive to 32 such pixels. Connected to and receiving the output of the shift register 42 is a similar shift register 44, again being a 32 s~age register. An accumulator 46 is . --11--.- -- : . - -~.~: -. - . . . :
~QS`~060 provided to receive and maintain an update~ a~gre~ate of the digitized pixel densities as the ~e are received from the A/D converter 18. As the microdensitometer is utilized in analyzing th~ ima~e, it is ~rocessed on S a pixel by pixel basis, such that the value within the accumulator 4h is updated by adding to the present value thereof the value of the new pixel within the ad~er 4R.
This updated value is then reduced by the value of the pixel 64 pixels away within the subtract circuit 5n. This value is then storea in the accumulator ~6. ~asically, the accumulator 46 contains an a~re~ate of the digitized values of 64 pixels, the value being updated by replacing the value of the pixel being scanned at any particular time for the value of the pixel scanned 6~ pixels earlier.
The output of the accumulator 46 is passed to the divider 52 to give an average of the density of the 6~
pixels stored therein. The divi~er 52 may, of course, be of any suitable nature as is well understoo~ ~y those skilled in ~he art. The pixel average output of the divider 52 is passed to the su~tract circuit 54 which al~o receives the output of the shift re~ister 42. It should be readily . ~ . .
~` apparent that this output is the di~iti~ed value of the ~` center plxel of that group o~ pixels representea hy the value stored within the accumulator 46. The OUtpllt of ~, 25 the subtract circuit 54 is thus the devi~tion of the density i of a pixel from the value of surroundin~ pixels on either side thereof. To a de~ree, the output of the element 54 ~; is thus a normalized den~ity output. This OlltpUt may be applied directly to the input of the full adder ~0 as shown ., `':
~ ~ .
"~ , .. .
1.0~ 0 in Fi~. 7. However, it is often times undesirAhle to have a negative input to the adder 2~ and hence the output of subtract circuit 54 is applied to an adder 5~ wherein the same is altered by a bias level. The bias applie~ to the adder 56 is most generally selected as the center of the operating range of pixel deviation such that positive and negative excursions about the bias level will be substantial-~y equal. The bias level virtually guarantees that the output of the adder 5~ will cause all inputs to the adder 20 to be positive.
It should be readily apparent from an understanding of the function of the circuitry of Fia. 7, that the hi~h frequency changes of density associatecl with flaws, cracks and the like will result in highly distinquishable de-viations above the average value of nei~hboring pixels as the same are contained in the accumulator ~. Thus, the normalized output resulting from the structure of Fig. 7 and combined with the offset techni~ue as shown in Fig. 5 will readily distinguish sharply var~ina density deviations above the background density.
Thus it can be seen that the objects of the invention have been satisfied by the techniques and structure presented hereinabove. ~Ihile in accordance with the patent statutes only the ~est mo~e and preferred embodiment of the invention has been presented and des-cribed in detail, it is to be understood that the invention is not limited thereto or thereby. Conse~uently, for an appreciation of the scope and breadth of the invention reference should be had to the followin~ claims.
'
Claims (10)
1. Apparatus for enhancing the data contained on a film transparency, characterized by:
first means for scanning said film transparency in discrete elements and digitizing the data of each of said elements;
a shifting circuit connected to said first means for shifting the digitized data of each and every element a predetermined number of positions toward the most significant bit of said data;
second means connected to said shifting circuit for converting said amplified digital data into analog signals; and third means connected to said second means for receiving said analog signals and recording the same on an output medium.
first means for scanning said film transparency in discrete elements and digitizing the data of each of said elements;
a shifting circuit connected to said first means for shifting the digitized data of each and every element a predetermined number of positions toward the most significant bit of said data;
second means connected to said shifting circuit for converting said amplified digital data into analog signals; and third means connected to said second means for receiving said analog signals and recording the same on an output medium.
2. The apparatus for enhancing data as recited in claim 1 which further includes an adder interconnected between the first means and said shifting circuit, said adder adding a fixed value to the digital value of each and every element.
3. The apparatus for enhancing data as recited in claim 1 which further includes a gradient removal cir-cuit interconnected between said first means and said shifting circuit; the gradient removal circuit normaliz-ing the digital value of the data of each of said elements with respect to certain surrounding elements.
4. The apparatus for enhancing data as recited in claim 1 wherein said first means comprises a micro-densitometer.
5. The apparatus as recited in claim 1 wherein said second means comprises a digital to analog converter and wherein said third means comprises a film recorder.
6. The method for enhancing the data contained on a film transparency, characterizing the steps of:
a) sectioning said transparency into specific discrete elements;
b) digitizing the data contained in each of said elements, said data being a function of the density of the elements;
c) amplifying the digitized data of each and every element by the same order of magnitude by shifting the digitized data a specific number of positions toward the most significant bit thereof;
d) converting the amplified digitized data of each of said elements to an analog data form; and e) recording the analog data on an output medium.
a) sectioning said transparency into specific discrete elements;
b) digitizing the data contained in each of said elements, said data being a function of the density of the elements;
c) amplifying the digitized data of each and every element by the same order of magnitude by shifting the digitized data a specific number of positions toward the most significant bit thereof;
d) converting the amplified digitized data of each of said elements to an analog data form; and e) recording the analog data on an output medium.
7. The method for enhancing data as recited in claim 6, which further includes the step of adding a preselected fixed value to the digitized value of the data of each and every element.
8. The method for enhancing data as recited in claim 7 which further includes the step of normalizing the value of the digitized data of each and every element with respect to the value of the digitized data of certain neighboring elements of each of said elements.
9. The method for enhancing data as recited in claim 6 which further includes the step of normalizing the value of digitized data of each and every element with respect to the digitized data of certain of the neighbor-ing elements of each of said elements.
10. The method for enhancing data as recited in claim 6 which further includes the step of adding a pre-selected fixed value to the digitized value of the data of each and every element.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US05/546,862 US3988602A (en) | 1975-02-03 | 1975-02-03 | Method and apparatus for enhancing data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA1056060A true CA1056060A (en) | 1979-06-05 |
Family
ID=24182345
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA243,103A Expired CA1056060A (en) | 1975-02-03 | 1976-01-07 | Method and apparatus for enhancing data |
Country Status (6)
| Country | Link |
|---|---|
| US (1) | US3988602A (en) |
| JP (1) | JPS51115818A (en) |
| CA (1) | CA1056060A (en) |
| DE (1) | DE2603533A1 (en) |
| FR (1) | FR2299766A1 (en) |
| GB (1) | GB1512891A (en) |
Families Citing this family (33)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4101961A (en) * | 1977-02-16 | 1978-07-18 | Nasa | Contour detector and data acquisition system for the left ventricular outline |
| JPS54121043A (en) * | 1978-03-13 | 1979-09-19 | Fuji Photo Film Co Ltd | X-ray photograph processing method |
| JPS6031389B2 (en) * | 1978-04-05 | 1985-07-22 | 日本電気株式会社 | Television video signal special effects equipment |
| US4213150A (en) * | 1978-04-21 | 1980-07-15 | Northrop Corporation | Real-time edge processing unit |
| DE2820965A1 (en) * | 1978-05-12 | 1979-11-15 | Agfa Gevaert Ag | PHOTOGRAPHIC COPY DEVICE |
| US4253154A (en) * | 1979-01-11 | 1981-02-24 | North American Philips Corporation | Line scan and X-ray map enhancement of SEM X-ray data |
| JPS55163472A (en) * | 1978-12-26 | 1980-12-19 | Fuji Photo Film Co Ltd | Radiant ray image processing method |
| JPS5588741A (en) * | 1978-12-26 | 1980-07-04 | Fuji Photo Film Co Ltd | Method of treating gradation of xxray picture of breast and its device |
| JPS55116340A (en) * | 1979-02-28 | 1980-09-06 | Fuji Photo Film Co Ltd | Method and device for processing gradation of radiation picture |
| JPS55116339A (en) * | 1979-02-28 | 1980-09-06 | Fuji Photo Film Co Ltd | Method and device for processing gradation of radiation picture of chest |
| CH642211A5 (en) * | 1979-03-08 | 1984-03-30 | Siemens Ag Albis | CORRECTION CIRCUIT TO IMPROVE THE SHARPNESS OF TELEVISION PICTURES. |
| US4265532A (en) * | 1979-06-21 | 1981-05-05 | Logetronics Inc. | Photo printing by intensity and velocity modulation |
| US4394737A (en) * | 1979-07-11 | 1983-07-19 | Fuji Photo Film Co., Ltd. | Method of processing radiographic image |
| US4375068A (en) * | 1980-04-21 | 1983-02-22 | Technicare Corporation | Radiographic apparatus and method with logarithmic video compression |
| US4345314A (en) * | 1980-11-19 | 1982-08-17 | International Business Machines Corporation | Dynamic threshold device |
| US4415980A (en) * | 1981-03-02 | 1983-11-15 | Lockheed Missiles & Space Co., Inc. | Automated radiographic inspection system |
| DE3149483A1 (en) * | 1981-12-14 | 1983-06-23 | Siemens AG, 1000 Berlin und 8000 München | X-RAY DIAGNOSTIC DEVICE FOR CREATING SUBTRACTION IMAGES |
| JPS598942A (en) * | 1982-07-09 | 1984-01-18 | 株式会社東芝 | X-ray image apparatus |
| JPS5928144A (en) * | 1982-08-09 | 1984-02-14 | Fuji Photo Film Co Ltd | Radiation picture reproducing device |
| JPS6096075A (en) * | 1983-10-31 | 1985-05-29 | Konishiroku Photo Ind Co Ltd | Picture reading method |
| JPS60154244A (en) * | 1984-01-24 | 1985-08-13 | Fuji Photo Film Co Ltd | Photographic image information detector |
| DE3527301A1 (en) * | 1984-07-31 | 1986-02-13 | Canon K.K., Tokio/Tokyo | Image reading device |
| US4829584A (en) * | 1985-03-26 | 1989-05-09 | Dainippon Screen Mfg. Co., Ltd. | Image input system |
| US4704633A (en) * | 1985-04-01 | 1987-11-03 | Fuji Photo Film Co., Ltd. | Method for reading out image information on an image having a wide dynamic range |
| JPS62116239A (en) * | 1985-11-15 | 1987-05-27 | Konishiroku Photo Ind Co Ltd | Method and apparatus for processing x-ray image |
| US4656525A (en) * | 1985-12-04 | 1987-04-07 | Polaroid Corporation | Electronic imaging copier |
| US4656524A (en) * | 1985-12-23 | 1987-04-07 | Polaroid Corporation | Electronic imaging copier |
| US4818649A (en) * | 1987-04-01 | 1989-04-04 | Sri International | Method for intensification and reflective read-out of underexposed film, radiographs, and the like |
| US4933754A (en) * | 1987-11-03 | 1990-06-12 | Ciba-Geigy Corporation | Method and apparatus for producing modified photographic prints |
| US4868651A (en) * | 1988-05-17 | 1989-09-19 | S&S Inficon, Inc. | Digital radiography with image brightness and contrast normalization |
| US5113422A (en) * | 1991-07-01 | 1992-05-12 | The United States Of America As Represented By The Secretary Of The Air Force | Radiographic interpretation trainer/test system |
| US5436979A (en) * | 1992-08-21 | 1995-07-25 | Eastman Kodak Company | Process for detecting and mapping dirt on the surface of a photographic element |
| US7542600B2 (en) * | 2004-10-21 | 2009-06-02 | Microsoft Corporation | Video image quality |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US3503689A (en) * | 1965-10-18 | 1970-03-31 | Technical Operations Inc | Microdensitometer |
| US3614311A (en) * | 1968-02-28 | 1971-10-19 | Hitachi Ltd | Apparatus for simultaneously displaying a plurality of images of an object being analyzed in an electron beam device |
| US3535443A (en) * | 1968-07-22 | 1970-10-20 | Gen Electric | X-ray image viewing apparatus |
| US3582651A (en) * | 1968-08-22 | 1971-06-01 | Westinghouse Electric Corp | X-ray image storage,reproduction and comparison system |
| JPS5116091B2 (en) * | 1971-09-03 | 1976-05-21 | ||
| US3813545A (en) * | 1973-04-12 | 1974-05-28 | Edax Int Inc | X-ray scan area mapping system |
-
1975
- 1975-02-03 US US05/546,862 patent/US3988602A/en not_active Expired - Lifetime
-
1976
- 1976-01-06 GB GB314/76A patent/GB1512891A/en not_active Expired
- 1976-01-07 CA CA243,103A patent/CA1056060A/en not_active Expired
- 1976-01-28 DE DE19762603533 patent/DE2603533A1/en not_active Withdrawn
- 1976-02-02 JP JP51009432A patent/JPS51115818A/en active Pending
- 1976-02-03 FR FR7602885A patent/FR2299766A1/en active Granted
Also Published As
| Publication number | Publication date |
|---|---|
| DE2603533A1 (en) | 1976-12-23 |
| FR2299766A1 (en) | 1976-08-27 |
| US3988602A (en) | 1976-10-26 |
| GB1512891A (en) | 1978-06-01 |
| JPS51115818A (en) | 1976-10-12 |
| FR2299766B1 (en) | 1980-08-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA1056060A (en) | Method and apparatus for enhancing data | |
| US5134573A (en) | Method to extend the linear range of images captured on film | |
| US5018024A (en) | Image processing apparatus using two-dimensional space frequency to discriminate image type | |
| EP0589376B1 (en) | Colour image reproduction of scenes with preferential tone mapping | |
| CN101778222B (en) | Image processing apparatus | |
| US4591923A (en) | Method and apparatus for intensifying sharpness of picture images | |
| KR850002390A (en) | X-ray diagnostic device | |
| US5659402A (en) | Image processing method and apparatus | |
| US4965679A (en) | Method for electronically duplicating film images while maintaining a high degree of image quality | |
| US5010398A (en) | Method for colour correction by dry dot etching using photographically produced mask | |
| US7453468B2 (en) | Intelligent color to texture converter | |
| JPS6173477A (en) | Picture processing method | |
| Rip | Digital recording and image processing of rock art by computer | |
| JP2902899B2 (en) | Automatic density correction method when reading radiation transmission test film | |
| JPS60136478A (en) | Picture processor | |
| JPS60191562A (en) | Picture processing unit | |
| Eizember et al. | Digital Image Analysis of NDT Radiographs | |
| JPS62256572A (en) | Picture processor | |
| JPS58114571A (en) | Picture recording method | |
| JPH077618A (en) | Method and device for processing color value | |
| Clark | Computer image processing of the Huntsville depot photograph | |
| Paluzzi et al. | Computer image processing in marine resource exploration | |
| JPH0574985B2 (en) | ||
| JPH05276387A (en) | Image reader | |
| Simonis | Scanner analyzer target |