WO2006090370A2

WO2006090370A2 - Enhancement of decompressed video

Info

Publication number: WO2006090370A2
Application number: PCT/IL2006/000229
Authority: WO
Inventors: Vitaly S. Sheraizin; Semion M. Sheraizin
Original assignee: Vlscom Limited
Priority date: 2005-02-22
Filing date: 2006-02-22
Publication date: 2006-08-31
Also published as: US7526142B2; WO2006090370A3; US7805019B2; US20060188168A1; US20090161754A1

Abstract

A method and apparatus for enhancing the video quality of compressed video signals adaptively removes distortions and ringing effects embedded in the decompressed images. The apparatus operates in conjunction with decoder devices installed in set-top boxes, satellite receivers, TV broadcast channel servers, digital still cameras, DVD players and recorders, large screen TV sets, media players, and the like.

Description

ENHANCEMENT OF DECOMPRESSED VIDEO FIELD OF THE INVENTION

[0001] The present invention relates generally to video signal decoding and more particularly to such decoding that can improve decompressed video signals.

BACKGROUND OF THE INVENTION

[0002] Video applications require that large amounts of data be transmitted at high bit rates and with a minimal amount of signal distortion. Since the available data bandwidth of a conventional transmission channel is limited, image coding techniques are utilized to compress the large amount of digital data to fit within the limited bandwidth.

[0003] Various video compression techniques are known in the art, such as those of the joint photographic expert group (JPEG), moving picture expert group (MPEG), such the MPEG-I, MPEG-2, and MPEG-4, H-compression, such as the H.261 H.262, H.263, and H.264, and the others. In most of these compression techniques, an image to be compressed is first divided into square blocks of pixels (e.g., an 8x8 pixel block). Each of these blocks is then transformed using discrete cosine transforms (DCT) into a transformed block (with 8x8-components) containing the DCT coefficients. These transformed blocks are then quantized (i.e. limited to one of a fixed set of possible values), and run-length encoded. Often, they are also variable length coded to further reduce the statistical redundancy present in the run-length coded data. A decoder on the receiving end of the transmission reconstructs the video stream from the transmitted, compressed signals.

[0004] As broadcast systems are required to provide an ever increasing amount of data utilizing the same data bandwidth, video signals are transmitted at lower and lower bit rates. For example, to increase the number of TV channels broadcasted to the viewers over a fixed data bandwidth, the bit rate of each channel is reduced to a rate between l.δMbps to 2.2Mbps. Unfortunately, transmitting data at too low a bit rate reduces the quality of the decompressed video stream. Furthermore, distortions are introduced into the decoded image, mainly consisting of annoying visual artifacts that are especially noticeable at medium and low bit rates. Distortions can be categorized into types, including "blocking" (or "blockness"), "blurring", and "wiggles", examples of which are shown in Figs. IA, IB and 1C, to which reference is now made. [0005] The blocking effect introduces artificial edges at the boundaries of the 8x8-pixel block, due to the quantization of the transform coefficients in each block. Fig. IA is an image of a man's face. Unfortunately, the coloring of his face in the area marked 10 is "blocky" rather than smooth. The edges of the blocks are perceived by the human eye as unnatural geometrical contours.

[0006] Quantization of transform coefficients also causes blurring of real contours present in the image, due to the reduction of the high frequency components in the DCT transformed blocks. In Fig. IB, the areas labeled 12 are blurred.

[0007] Distortion has another side effect, where some retained frequency components remain unbalanced, causing ripples near edges. These ripples, known as "wiggles" or "mosquito noise", cause those areas with high frequency components to appear, move and disappear at random points of the frame. This can be seen in Fig. 1C, in the areas labeled 14.

[0008] Decompressed video signals may include other noise effects as well, such as blotches and ringing.

BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The subject matter regarded as the invention is particularly pointed out and distinctly claimed in the concluding portion of the specification. The invention, however, both as to organization and method of operation, together with objects, features, and advantages thereof, may best be understood by reference to the following detailed description when read with the accompanying drawings in which:

[0010] Figs. IA, IB and 1C are exemplary decompressed images with exemplary types of distortions;

[0011] Fig. 2 is a block diagram illustration of an apparatus for improved picture quality, constructed and operative in accordance with the present invention;

[0012] Fig. 3 is a block diagram illustration of an exemplary distortion estimator, forming part of the apparatus of Fig. 2;

[0013] Fig. 4 is a detailed block diagram illustration of the distortion estimator of Fig. 3;

[0014] Fig. 5 is a block diagram illustration of an exemplary high contrast details analyzer forming part of the apparatus of Fig. 2;

[0015] Figs. 6A and 6B are graphical illustrations of the response of limiters, useful in understanding the operation of the character analyzer of Fig. 5;

[0016] Fig. 7 is a block diagram illustration of an exemplary parameter estimator forming part of the apparatus of Fig. 2;

[0017] Fig. 8 is a block diagram illustration of a controller forming part of the apparatus of Fig. 2;

[0018] Fig. 9 is a block diagram illustration of an exemplary adaptive temporal processor forming part of the apparatus of Fig. 2;

[0019] Fig. 10 is a block diagram illustration of an alternative embodiment of adaptive temporal processor, forming part of the apparatus of Fig. 2 and using an HR filter;

[0020] Fig. 11 is a block diagram illustration of an exemplary adaptive spatial processor forming part of the apparatus of Fig. 2; [0021] Figs. 12A and 12B are graphical illustrations of the response of limiters, useful in understanding the operation of the processors of Figs. 9 and 11;

[0022] Fig. 13 is a graphical illustration of a K_v function, useful in understanding the operation of the processor of Fig. 11; and

[0023] Fig. 14 is a flow chart illustration of a method of operating the apparatus of Fig. 2.

[0024] It will be appreciated that for simplicity and clarity of illustration, elements shown in the figures have not necessarily been drawn to scale. For example, the dimensions of some of the elements may be exaggerated relative to other elements for clarity. Further, where considered appropriate, reference numerals may be repeated among the figures to indicate corresponding or analogous elements.

DETAILED DESCRIPTION OF THE PRESENT INVENTION

[0025] In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be understood by those skilled in the art that the present invention may be practiced without these specific details. In other instances, well-known methods, procedures, and components have not been described in detail so as not to obscure the present invention.

[0026] Unless specifically stated otherwise, as apparent from the following discussions, it is appreciated that, throughout the specification, discussions utilizing terms such as "processing," "computing," "calculating," "determining," or the like, refer to the action and/or processes of a computer, computing system, or similar electronic computing device that manipulates and/or transforms data represented as physical, such as electronic, quantities within the computing system's registers and/or memories into other data similarly represented as physical quantities within the computing system's memories, registers or other such information storage, transmission or display devices.

[0027] Reference is now made to Fig. 2, which illustrates an exemplary and non-limiting block diagram of an apparatus 100 to improve the visual quality of images decompressed from compressed video signals, constructed and operative in accordance with the present invention. Apparatus 100 may operate in conjunction with various decoder devices 102, such as those installed in set-top boxes, satellite receivers, TV broadcast channel servers, digital still cameras, DVD players and recorders, large screen TV sets, media players, and the like, and may attempt to remove distortions and ringing effects found in the decompressed video signal, labeled Y, produced by such decoder devices 102.

[0028] Decompressed video signal Y may be decompressed from a signal encoded with coding techniques including, but not limited to, those of joint photographic expert group (JPEG), moving picture expert group (MPEG), such the MPEG-I, MPEG-2, and MPEG-4, h- compression, such as H.261, H.262, H.263 and H.264, Windows Media (WM) 9, and others. Such coding techniques typically represent the video signals with motion vectors and residual transform DCT coefficients. Each frame of the video signal is coded based on a prediction from one or more previously coded frames, and thus, properly decoding one frame requires first decoding one or more other frames. This temporal dependence between frames severely complicates a number of spatial and temporal processing techniques, such as translation, downscaling, and splicing.

[0029] In accordance with the present invention, apparatus 100 may comprise an adaptive temporal processor 110, an adaptive spatial processor 120, an estimation unit 130, and a controller 140. As described in more detail hereinbelow, estimation unit 130 may receive decompressed video signal Y and may analyze it to determine the extent of improvement necessary. Estimation unit 130 may include a distortion estimator 210, a parameter estimator 220, and a high contrast details analyzer 230. Distortion estimator 210 may estimate the amount of distortion DR by checking each block (e.g., 8x8-pixel) in an image.

[0030] Parameter estimator 220 may estimate the image complexity and may generate a value NC indicating the extent to which the image has changed from a previous image. The more complex an image, the more distorted the image may appear to the human eye. Parameter estimator 220 may also generate an hy value indicating a change in intensity between a pixel (ij) in two consecutive frames.

[0031] High contrast details analyzer 230 may analyze high contrast details in the image^ such as edges and characters. High contrast details analyzer 230 may generate a per-pixel value CH_y indicating the extent of high contrast details at each pixel in the frame.

[0032] Controller 140 may convert the parameters generated by estimation unit 130 into temporal control parameters, which may control aspects of the processing of adaptive temporal processor 110, and spatial-temporal control parameters, useful for controlling adaptive spatial processor 120.

[0033] Adaptive temporal processor 110 may operate on decompressed signal Y, while adaptive spatial processor 120 may operate on the output, a signal Y_IP, of processor 110. The result may be an improved decompressed signal Yp. Processors 110 and 120 may include temporal and spatial components such as are known in the art. In addition, processors 110 and 120 may include additional components controlled by the spatial and temporal control parameters received from controller 140. The latter may attempt to reduce the distortions and other low quality effects found in decompressed video signals. The output of adaptive spatial processor 120, the signal Yp, may be an improved video signal with reduced distortions. [0034] Reference is now made to Fig. 3, which shows a non-limiting block diagram of distortion estimator 210 operative in accordance with an exemplary embodiment of the present invention. Distortion estimator 210 may include a vertical high pass filter (HPF) 310, a horizontal low pass filter (LPF) 320, an into line integrator 330, a line-to-line periodicity detector 340, and an averager 350.

[0035] Distortion estimator 210 may measure the blockness in the frame by counting the high frequency components along the edges of each k by k pixel block of the frame. To do so, vertical HPF 310 may transfer only the high frequency components Y_HF of decompressed signal Y. Horizontal LPF 320 may reduce noise in the high frequency component signal Y_HF, generating thereby a signal Ϋ_HF. Into line integrator 330 may sum, for each line in the frame, the number of Ϋ_HF components every A* pixel in the line. Line-to-line periodicity detector 340 may sum the number of high frequency components in each k^ft line and may generate the result, a value DR', as described in detail hereinbelow. The value DR' may indicate the amount of distortion in the current frame. Averager 350 may generate the distortion value DR by averaging DR' with the distortion value DR computed for the previous frames.

[0036] Fig. 4, to which reference is now made, shows an exemplary functional block diagram of line integrator 330 and line-to-line detector 340 for a block size of 8 by 8 pixels. Other block sizes are possible and are incorporated in the present invention.

[0037] Into line integrator 330 may sum the high frequency components Ϋ_HF of every 8^th pixel along a line / of the frame. The summation may be performed using the following equation:

AM

*. =£ iV(so (l)

where N is the number of pixels in a line.

[0038] To find column edges, line-to-line periodicity detector 340 may sum the output of line integrator 330 every 8^th line, starting at different lines. Detector 340 may then sum those blocks which are distorted, defined here as those outputs above an average value.

[0039] Detector 340 may include eight summation units 341-1 through 341-8, eight adders 342-1 through 342-8, eight operators 343-1 through 343-8, an averager 410 and a summer 420. Each summation unit 341-i may accumulate the high frequency components of each 8^th line starting from the i^til line. For example, summation unit 341-1 may operate on lines 1, 9, 17, ..., M-I₃ while summation unit 341-2 may operate on lines 2, 10, 18, ..., M-I, and so on. Mathematically, the i^fll summation unit 341-i may operate as follows:

where M is the maximum number of lines in a frame (e.g. M=480 for NTSC). The index i varies from 1 to 8.

[0040] Adders 342-i and operators 343-i together may operate to provide the value Si only if it is above an average value S of the set of values Sj, where average value S may be computed by averager 410 as follows:

[0041] Each adder 342-i may subtract average value S from its associated Si and may produce a difference value Δj.

[0042] An operator 343-i may utilize its associated difference value Δj to generate a signal Vi which has the value of Si only if difference value Δ, is positive, as follows:

_y J° ^ ^≤0 (4)

[0043] The values yj may indicate the extent of distortion for each i^Λ portion of the frame. Summer 420 may accumulate the values of yi to generate distortion level DR' for the current frame.

[0044] Reference is now made to Fig. 5, which illustrates a non-limiting and exemplary block diagram of high contrast details analyzer 230. High contrast objects in video signals (e.g., text characters) may be characterized by high frequency components. In accordance with a preferred embodiment of the present invention, they may be processed as distortions. To detect high contrast details, high contrast details analyzer 230 may comprise a high pass filter (HPF) 510, an adder 520, a limiter 530, a two-dimensional low-pass filter (LPF) 540 and a second limiter 550. [0045] HPF 510 may generate high frequency components XHF from input decompressed video signal Y. Adder 520 may subtract a first threshold THDl from the X_HF components, thereby producing a shifted signal X_nI. THDl may be defined by the expected intensity levels of high contrast details. For example, THDl may be set to THDl=0.25 Y_1113x, where Y_103x may be the maximum possible intensity level for the video signal (e.g., 256).

[0046] Limiter 530 may limit the signal intensities of the output of adder 520 (a signal X^) to those below a given threshold THD2, where THD2 may be set, for example, to THD2=0.1 Y_n^. Specifically, the signal X₀₂ output by limiter 530 may be written mathematically as follows:

[0047] 2D-LPF 540 may filter the signal X_n2 to detect points (dots) in decompressed frame Y. LPF 540 may have a cutoff point which may be a function of both an expected contrast level and the size of the smallest details.

[0048] To indicate the presence of text characters in decompressed frame Y, limiter 550 may limit the intensities of signal X_n3, generated by LPF 540, to those below a given threshold THD3, where THD3 may be set to THDS=O-OSY_n^_x, thereby generating a per-pixel CHy signal. The level of per-pixel signal CHy maybe defined mathematically as follows:

CH _u = \ Xm ifθ < Xm ≤ THm ₍₆₎

[THD3 if X_n3 > THD3

[0049] Figs. 6A and 6B, to which reference is now briefly made, are graphical illustrations of the response of limiters 530 and 550 respectively.

[0050] Reference is now made to Fig. 7, a non-limiting block diagram of parameter estimator 220, constructed and operative in accordance with an exemplary embodiment of the present invention. As mentioned above, parameter estimator 220 may estimate the image complexity NC and may generate signal change values hg. Image complexity NC may indicate the extent to which the image has changed from a previous frame. The signal change values hy may indicate, per pixel (ij), the intensity change between two consecutive frames. [0051] To generate the signal change values h_y, parameter estimator 220 may take the difference of intensity between consecutive frames, for each pixel. For this purpose, parameter estimator 220 may comprise a frame memory 710 and an adder 720. Frame memory 710 may delay decompressed frame Y(n), thereby producing a previous frame Y(n-1). Adder 720 may generate a difference frame Δ_F between previous frame Y(n-1) and current input frame Y(n).

[0052] Parameter estimator 220 may then smooth difference frame Δ_F with an absolute value operator 730 and a horizontal low pass filter (LPF) 740. Absolute value operator 730 may take the absolute value for each pixel in difference frame Δ_F, generating a signal | Δp | , and horizontal LPF 740 may generally reduce any noise that maybe present in the signal | Δ_F | . The intensity of each pixel (i, j) in signal | Δ_F | may be the value %

[0053] To generate image complexity NC, parameter estimator 220 may comprise a frame intensity change generator 745, which may utilize the pixel intensity changes hy, a histogram difference generator 775, a histogram normalizer 780 and a weighted summer 790.

[0054] Histogram difference generator 775 may determine how different a histogram of the intensities Y of the current frame (n) is from that of a previous frame (n-1). An image of the same scene generally may have a very similar collection of intensities, even if the objects in the scene have moved around, while an image of a different scene may have a different histogram of intensities. Thus, histogram difference estimator 775 may measure the extent of change in the histogram.

[0055] Histogram difference generator 775 may comprise a histogram estimator 770, a histogram storage unit 715 and an adder 725. Adder 725 may generate a difference histograms Δmst by taking the difference between the histogram for the current frame (n) as provided by histogram estimator 770 and that of the previous frame stored in histogram storage unit 715.

[0056] Histogram normalizer 780 may normalize difference histogram A_Sist as follows:

A = ^BU /_? Λ

H-St _N *_M V t

where N and M are the maximum number of lines and columns of the frame, respectively.

[0057] Frame intensity change generator 745 may determine an overall change value V_D indicating the extent of significant change in the frame and may comprise a summation unit 750 and a normalizer 760. Summation unit 750 may sum the values of h_g for all pixels in the in signal I ΔF I . Mathematically, summation unit 750 may perform the following equation:

[0058] Normalizer 760 may normalize frame intensity change V_D, by the frame size and the maximum intensity levels. For example, normalizer 760 may divide frame intensity change V_D by

[0059] Weighted summer 790 may generate image complexity value NC from a weighted sum of normalized frame intensity change V_D and normalized difference histogram A^_ist , as follows:

NC = K_HA_Wjst +K_vV_D (7c)

where KH and Kv are the weighting coefficients and each may have a value between 0 and 1.

[0060] Reference is now made to Fig. 8, which illustrates controller 140 in accordance with an exemplary embodiment of the present invention. Controller 140 may generate temporal control parameters for adaptive temporal processor 110 and spatial-temporal control paraitieters for adaptive spatial processor 120 from the estimation and prediction parameters DR, CHy, hy and NC discussed hereinabove.

[0061] Controller 140 may generate a temporal threshold THDip to define the level of low contrast small details in the frame. Typically, in low contrast images, the human eye can detect small details in an image only if their intensity levels are 3 times higher than the average noise (i.e., distortion). From this, the temporal threshold THDipis defined, as follows:

where DR^* is a normalized value of DR, normalized by the quantity of blocks in a frame (e.g. N*M/8*8 ).

[0062] Controller 140 may generate a low-frequency coefficient K_F which may be used to attenuate low frequency components to reduce blockness in the image.

K_F= l-0.5DR^* (9) [0063] Controller 140 may generate a texture coefficient

and a contrast coefficient K¹¹COiIt which may be used to reduce noise in high contrast and low contrast (i.e., texture) signals, respectively.

[0064] Controller 140 may generate a spatial-temporal threshold THD_SP to define the level of low contrast, small details in the frame and a spatial text coefficient K^sp _text to adjust the texture sharpness in a video signal, as follows:

THD_SP ^ 3(l+DR^*) (12)

KZ = K^M-DIt) (13)

[0065] where

and may be equal or greater than 1.

[0066] Controller 140 may generate a per-pixel, spatial contrast coefficient

to adjust the sharpness of the small details, as follows:

κz_t(i, J)

(i4)

where K^_nto is a maximum contrast coefficient and may be greater than 1 and CHy is a normalized value of per-pixel signal CHy, normalized by CHm_3x, a maximum possible value of the

[0067] Controller 140 may generate a per-pixel recursion coefficient K^ij) to reduce noise, as follows:

K_rec(iJ) = K_ιeco[l + 0.25(DR^* +CHj)I O⁵)

where K_1x-0 is a maximum recursion coefficient and may be equal or smaller than 0.5.

[0068] As can be seen, the values OfK⁸¹Ut, Kn₅₀(Lj), and K^sp _cont(Lj) depend on the amount of noise (CH) and distortion (DR) in the image. High values of these coefficients may imply high noise and distortion in the image.

[0069] Controller 140 may generate a ringing coefficient KdHg(Lj) per pixel (Lj), to eliminate ringing effects, as follows: K_ring(iJ) = l- 0.5h; (16)

where hj/ is the normalized value of hy, normalized by hij_max. The ringing coefficient Krin₈(ij) may be used to reduce or eliminate ringing effects as will be described in greater detail hereinbelow.

[0070] Reference is now made to Fig. 9, which shows a non-limiting block diagram of adaptive temporal processor 110, constructed and operative in accordance with an exemplary embodiment of the present invention. Adaptive temporal processor 110 may comprise a standard temporal processor 800, a texture improver 810, a small details sharpener 816 and a blockness reducer 830.

[0071] Standard temporal processor 800 may comprise a temporal high pass filter (HPF) 811, a temporal low pass filter (LPF) 812, a two-dimensional (2D) spatial LPF 813, a vertical HPF 814, and a horizontal HPF 815 and may operate to temporally process decompressed frame Y.

[0072] Texture improver 810 may attempt to sharpen textual elements, if present, in decompressed frame Y and may comprise limiters 820-1 and 820-2, a horizontal LPF 860, a vertical LPF 870, a texture sharpener 850 and an adder 880. A graphical representation of both limiters 820-1 and 820-2 is provided in Fig. 12A

[0073] Texture improver 810 may operate on the vertical and horizontal high frequency components (Le. the components along a column and a row, respectively) of frame Y, generated by vertical HPF 814 and horizontal HPF 815, respectively. To sharpen text and other textured items and to reduce distortions without affecting the image quality, limiters 820 may limit the intensities of high frequency signals X_HF to below THDxp.

[0074] Low pass filters 860 and 870 may reduce noise from the output signals Ytestv and Ytexth of limiters 820-1 and 820-2, respectively. Specifically, low pass filter 860 may be a horizontal LPF operating on vertical signal Yt_6xLv and, similarly, low pass filter 870 may be a vertical LPF operating on horizontal signal Y_texth- It will be appreciated that such a structure may reduce noise without affecting the image sharpness. Experiments show that the expected noise reduction is around 6 dB.

[0075] Adder 880 may sum the thresholded and noise reduced signals to generate a texture signal Ytø& and texture sharpener 850 may adjust the sharpness of texture signal Y_text with texture coefficient It will be appreciated that sharpener 850 may reduce the texture sharpness with low

(present when the distortion level DR is high, as per equation 10) and may increase the sharpness for high values of K

^_ta* The values may be determined according to Eq. (10).

[0076] Small details sharpener 816 may comprise a contrast sharpener 840 and three adders 842, 844 and 846. Adders 842 and 844 may sum together the inputs and outputs of limiters 820, generating signals with strongly contrasting small details. Adder 846 may produce a linear sum of the outputs of adders 842 and 844 to generate a small details signal Ycnt- Contrast sharpener 840 may adjust the sharpness of the small details in Y_∞nt using the contrast coefficient

determined in accordance with Eq. (11). Contrast sharpener 840 may be a linear operator, i.e., it may reduce the small details sharpness with low values

and may increase the sharpness with high values of K^

[0077] Distortion reducer 830 may be a linear operator and may operate on the low pass filtered signals Y_LP generated by 2D spatial LPF 813. Distortion reducer 830 may attenuate the low frequency components with coefficient K^, in order to reduce blockness in the frame. The blockness coefficient K_F may be determined in accordance with Eq. (9).

[0078] A summation unit 890 may accumulate the outputs of temporal low pass filter 812 (a noise reduced version of decompressed image Y), texture improver 810, small details sharpener 816 and distortion reducer 830 and may provide adaptive spatial processor 120 with signal Yπ>.

[0079] In accordance with an alternative embodiment of the present invention, adaptive temporal processor 110 may comprise an adaptive recursive filter or infinite impulse response (HR) filter. This may allow a reduction in the number of frame memories in the system.

[0080] A non-limiting block diagram of alternative adaptive temporal processor, labeled 110', is shown in Fig. 10. Processor 110 may have an adaptive recursion coefficient K_r which may be changed according to the changes in intensity between the input Y and the processed output Yrp and as a function of the distortion level DR.

[0081] Processor 110' may comprise adders 822 and 823, a frame memory 824, a weighting unit 825, a low pass filter 826, limiters 827, a normalizer 828, an absolute value operator 829, and a recursion operator 831. [0082] Adder 823 may generate a difference frame Δ between the previous output Yπ>, stored in frame memory 824 and current input frame Y.

[0083] Low pass filter LPF 826 may smooth difference frame Δ and absolute value operator 829 may generate the absolute value | Δ* | of the output of low pass filter 826. Limiter 827 may limit the value| Δ* | below a threshold level THD, described hereinbelow, and generate a signal | Δ* di_m. Normalizer 828 may then normalize the signal| Δ* |κ_m by THD, thereby producing a value between 0 and 1.

[0084] The THD level may be set to an average block signal value and may be computed as follows:

THD = THD₀ * (1 + pDR*) (17) where THD₀ = 0.01 * Y_110x and p may be equal to 2, 3, or 5.

[0085] Recursion operator 831 may generate adaptive recursion coefficient K_r for the filter as follows:

K₁=K^ (I - I Δ k/ THD) (18) where K₁₀ may vary between 0 and 1.

[0086] Weighting unit 825 and adder 822 may implement the IIR filter by multiplying the difference frame Δ by adaptive recursion coefficient K₁ and subtracting the weighted result from input frame Y.

[0087] It will be appreciated by a person skilled in the art that for high probability blocks (Le. blocks with low and mid contrast levels of blockness), K_r is high and the noise and blockness reduction is relatively efficient.

[0088] Reference is now made to Fig. 11, which shows a non-limiting block diagram of adaptive spatial processor 120, constructed and operative in accordance with an exemplary embodiment of the present invention. Adaptive spatial processor 120 may comprise a standard spatial-temporal processor 900, a ringing eliminator 910, a texture improver 914 and a small details sharpener 916.

[0089] Spatial-temporal processor 900 may perform spatial processing on signal Yπ> provided by processor 110 or processor 110'. Spatial-temporal processor 900 may comprise a two-dimensional (2D) spatial low pass filter (LPF) 911, a vertical high pass filter (HPF) 912, and a horizontal low pass filter (LPF) 913.

[0090] Ringing eliminator 910 may attempt to remove ringing effects in the decompressed video signal. A ringing effect (or corona effect) may appear around the edge of the decompressed block and may cause noise in the high frequency component of the decoded video signal. The ringing effect may induce annoying visual artifacts that are especially noticeable on large screen TVs.

[0091] Ringing eliminator 910 may comprise limiters 921 and 922, anti-ringing units 941 and 942 as well as adders 981 and 982. A graphical representation of both limiters 921 and 922 is provided in Fig. 12B.

[0092] Ringing eliminator 910 may operate on the vertical and horizontal high frequency components of signal Yπ>, generated by vertical HPF 912 and horizontal HPF 913, respectively. To determine a texture level for horizontal and vertical components in signal Yrp, limiters 921 and 922 may limit the intensities of high frequency signals XHF.V and XHFJH, respectively, to below THDsp. Adders 942 and 944 may sum together the inputs and outputs of limiters 921 and 922, respectively, generating signals, labeled V^v and Vi_nJ1, respectively, with strongly contrasting small details.

[0093] Anti-ringing units 941 and 942 may receive the Vj_n signals and may attenuate them, using an attenuation graph, such as that shown in Fig. 13, to which reference is now briefly made.

[0094] Fig. 13 graphs a fixed coefficient K_v as a linear inverse function 999 of the input signal Vj_n. Anti-ringing units 941 and 942 may then weight fixed coefficient K_v by per-pixel, ringing coefficient Kππ/ij) to generate an anti-ringing attenuation coefficient K₃₁, as follows:

K_ar = K_ring(ij)*K_v (19)

[0095] Texture improver 914 may attempt to sharpen textual elements, if present, in the outputs of limiters 921 and 922. Texture improver 914 may comprise an adder 983, a temporal LPF 950 and a texture sharpener 970. Adder 983 may generate a texture signal Y_text and temporal LPF 950 may operate on textural signal Y_text to reduce noise using per-pixel recursion coefficient K₁Ec(Ij), which may be determined in accordance with Eq. (15). Temporal LPF 950 may be an infinite impulse response (DR) filter and may utilize per-pixel recursion coefficient Kreo(i j) as its recursion coefficient. It will be appreciated that recursion coefficient K^i ,j) may be a function of the distortion level and small details in the image. The level of noise reduction is higher for higher values of K₁J(Ij).

[0096] Texture sharpener 970 may be a linear operator and may adjust the texture sharpness of the output of temporal LPF 950 with the value of K⁸¹UHt determined in accordance with Eq. (13).

[0097] Small details sharpener 916 may comprise an adder 984 and a contrast sharpener 960. Adder 984 may sum the outputs of anti-ringing units 941 and 942, generating a signal, labeled Y_∞nt, with strongly contrasting, small details. Contrast sharpener 960 may adjust the sharpness of the small details based on the per-pixel values of K^SPcont(ij) determined in accordance with Eq. (14). Contrast sharpener 960 may be a linear operator, i.e., it may reduce the small details sharpness with low values of K^SPc_<mt(i j) and may increase the sharpness with high values of

K^SPcont(ij).

[0098] A summation unit 990 may accumulate the outputs of 2D spatial LPF 911, texture improver 914 and small details sharpener 916 to generate improved quality video signal Yp.

[0099] It should be appreciated by a person skilled in the art that the components of apparatus 100 disclosed herein may be hardware components, software components, firmware components, or any combination thereof.

[00100] Reference is now made to Fig. 14, which shows a non-limiting flowchart 1400 describing the operation of apparatus 100 in accordance with an exemplary embodiment of the present invention. Apparatus 100 may perform the steps of Fig. 14 in order or in parallel.

[00101] At step S1410, apparatus 100 may receive decompressed signal Y. At step S1420, estimation unit 130 may estimate the amount of distortion DR, the image complexity NC, the intensity change value hjj, and the extent of high contrast details CH, as described in greater detail above. At step S 1430, controller 140 may generate the temporal and spatial-temporal control parameters according to the equations provided hereinabove.

[00102] Utilizing the temporal control parameters, at step S 1440, adaptive temporal processor 110 may temporally process decompressed signal Y to eliminate temporal dependencies and to reduce noise and blockness. In addition, adaptive temporal processor 110 may adjust the texture sharpness and the small details sharpness. At step S 1450, adaptive spatial processor 120 may spatially process the output of adaptive temporal processor 110. The spatial processing may involve reducing noise and ringing effects as well as adjusting the sharpness of small details and texture components. Adaptive spatial processor 120 may utilize the spatial-temporal control parameters. At step S 1460, adaptive spatial processor 120 may output the signal Y_P which has enhanced video quality.

[00103] While certain features of the invention have been illustrated and described herein, many modifications, substitutions, changes, and equivalents will now occur to those of ordinary skill in the art. It is, therefore, to be understood that the appended claims are intended to cover all such modifications and changes as fall within the true spirit of the invention.

Claims

CLAIMSWhat is claimed is:

1. An apparatus comprising: an estimation unit to generate visual quality parameters that indicate the visual quality of decompressed video frames; and a decompressed video enhancer to substantially improve the visual quality of said decompressed video frames using their associated parameters.

2. The apparatus of claim 1 wherein said visual quality is defined at least as the extent of distortions and ringing effects.

3. The apparatus of claim 1, wherein said apparatus is installed in at least one of: a set-top box, a satellite receiver, a television (TV) broadcast channel server, and a large screen TV set.

4. The apparatus of claim 1, wherein said apparatus is installed in at least one of: a digital still camera and a mobile video.

5. The apparatus of claim 1, wherein said apparatus is installed in at least one of: a digital videodisc (DVD) player and a DVD recorder.

6. The apparatus of claim 1, wherein said apparatus is installed in a media player.

7. The apparatus of claim 1, wherein said decompressed video frames are decompressed according to a decoding technique including at least one of: a joint photographic expert group (JPEG), a moving picture expert group (MPEG), an H- compression, and a Windows Media (WM).

8. The apparatus of claim 1, wherein said estimation unit comprises: a distortion estimator to estimate a distortion level DR in a frame; a parameter estimator to estimate an image complexity value NC and a per pixel intensity change hy-; and a high contrast details analyzer to estimate high contrast small details levels CHi j in a frame.

9. The apparatus of claim 8, wherein said distortion estimator comprises a blockness level determiner.

10. The apparatus of claim 9, wherein said blockness level determiner comprises: an into line integrator to sum high frequency components in every k pixels along a line of said frame; and a line-to-line periodicity detector to sum the output of said line integrator every k^a line, starting at different lines and to sum those blocks which are distorted.

11. The apparatus of claim 8, wherein said parameter estimator comprises: an intensity change generator to generate per-pixel intensity change hy; and an image complexity generator to determine per-frame image complexity level

NC.

12. The apparatus according to claim 11 and wherein said intensity change generator comprises: a first frame memory to generate a difference frame Δ_F of said decompressed video frame and a previous decompressed video frame; an absolute value unit to take an absolute value of said difference frame Δ_F; and a horizontal low pass filter (LPF) to reduce noise in the output of said absolute value unit.

13. The apparatus according to claim 11 and wherein said image complexity generator comprises: a histogram difference generator to determine how different a histogram of the intensities of said frame is from that of a previous frame; a frame intensity change generator to determine an overall change value from said per-pixel intensity changes hy, said overall change value indicating the extent of significant change in said frame; and an image complexity generator to generate said per-frame image complexity level NC from a weighted sum of the outputs of said generators.

14. The apparatus of claim 8, wherein said high contrast details analyzer comprises: a shifter to shift said frequency components of said decompressed video frame downward by a first threshold, wherein said first threshold THDl is defined by expected intensity levels of high contrast details; a first limiter to limit high frequency components of said decompressed video frame to an intensity level below a second threshold; a two-dimensional low pass filter operating on the output of said limiter to detect dots in said frame; and a second limiter to limit the output of said two-dimensional low pass filter to an intensity level below a third threshold.

15. The apparatus of claim 8 wherein said enhancer comprises: an adaptive temporal processor to adapt temporal processing of each said decompressed video frame with at least one of said associated parameters; and an adaptive spatial processor to adapt spatial processing of the output of said adaptive temporal processor with at least one of said associated parameters.

16. The apparatus of claim 15 and further comprising a controller to generate at least temporal control parameters for said adaptive temporal processor and spatial-temporal control parameters for said adaptive spatial processor from at least one of said visual quality parameters.

17. The apparatus of claim 16, wherein said temporal control parameters comprise at least one of: a temporal threshold THD_TP determined by:

a temporal texture coefficient

K^ = 1 - DR^* ; a temporal contrast coefficient

= 1 -DR^* ; and a low frequency coefficient K_F determined by: K_F= 1 - 0.5DR^* , wherein said DR^* is a normalized value of said distortion level DR.

18. The apparatus of claim 16, wherein said spatial-temporal control parameters comprise at least one of: a spatial-temporal threshold THDsp determined by: THD_8P = 3(1+Z⁾R^*) , a per-pixel, recursion coefficient Kr_ec(i,j) determined by: KJ$_t j) = K_reco [l + 02S(DR^* + C#_r/)]; a per-pixel ringing coefficient Kπn_g(i,j) determined by: K_nng(i, J)= I- 0.5/ϊ_y ^* , a spatial-temporal texture coefficient K^sp _text determined by:

K^ = K_{text 0}(l- DR^*) ; and a per-pixel, spatial-temporal contrast coefficient K^sp _cont(i,j)

KL(JJ) = K_αonto[l- 0.5(DR^* -CHj)I wherein said DR^* is a normalized value of said distortion level DR, said CHy^* are normalized values of said CH_y values and said hy are normalized values of said hy, and where said K₀₀,,^ is a maximum contrast coefficient, said K_rec.o is a maximum recursion coefficient, and said Ktexto is a maximum texture coefficient.

19. The apparatus of claim 18, wherein said per-pixel, ringing coefficient Kring(ij) is used to reduce ringing effects in said decompressed frame.

20. The apparatus of claim 17, wherein said adaptive temporal processor comprises: a temporal processor to temporally process said decompressed video frame; a texture improver to receive vertical and horizontal high pass filtered signals from said temporal processor and to attempt to sharpen textual elements in said frame therefrom: a small details sharpener to adjust the sharpness of small details in said frame; and a distortion reducer to reduce blockness in said frame by attenuating low frequency components of said frame received from said temporal processor .

21. The apparatus of claim 20 and wherein said texture improver comprises: a first limiter to limit an input signal level below said temporal threshold THDχp₅ wherein the input signal includes vertical high frequency components of said decompressed frame; a second limiter to limit an input signal level below said temporal threshold THDTP, wherein the input signal includes horizontal high frequency components of said decompressed frame; a horizontal low pass filter (LPF) to pass low frequency components of a signal produced by said first limiter; a vertical LPF to pass low frequency components of a signal produced by said second limiter; and a texture sharpener to adjust texture sharpness of the sum of the outputs of said LPFs according to said temporal texture coefficient K¹¹W

22. The apparatus of claim 21 and wherein said small details sharpener comprises: a small details unit to generate a small details signal from said vertical and horizontal high frequency components and the outputs of said limiters; and a small details sharpener to adjust a sharpness of said small details signal according to said temporal contrast coefficient

23. The apparatus of claim 21 and wherein said distortion reducer comprises a blockness sharpener to reduce the level of low frequency components of said decompressed video frame according to said blockness coefficient K_F.

24. The apparatus of claim 15, wherein said adaptive temporal processor comprises an infinite impulse response (DR) filter using an adaptive recursion coefficient determined at least from said estimated distortion level DR.

25. The apparatus of claim 18, wherein said adaptive spatial processor comprises: a spatial-temporal processor to perform spatial processing on said decompressed video frame; a texture improver to adjust textual elements in said frame; a small details sharpener to adjust the sharpness of small details in said frame; and a ringing reducer to receive vertical and horizontal high pass filtered signals from said spatial-temporal processor and to reduce ringing effects in said frame.

26. The apparatus of claim 25, wherein said ringing reducer comprises: a first limiter to limit an input signal level below said spatial threshold THDsp, wherein the input signal includes vertical high frequency components of said decompressed frame; a second limiter to limit an input signal level below said spatial threshold THD_SP, wherein the input signal includes horizontal high frequency components of said decompressed frame; a vertical anti-ringing unit to reduce ringing effects in said vertical high pass filtered signals according to said ringing coefficient K^y₎; and a horizontal anti-ringing to reduce ringing effects in said vertical high pass filtered signals according to said ringing coefficient Kήn_gC_j).

27. The apparatus of claim 25 wherein said small details sharpener comprises a high contrast sharpener to adjust a sharpness of said small details signal according to said spatial- temporal contrast coefficient K^sp _cont(y).

28. The apparatus of claim 25 and wherein said texture improver comprises: a temporal low pass filter (LPF) to reduce noise according to said recursion coefficient K_reo; and a texture sharpener to adjust texture sharpness of an output of said temporal LPF according to a value of said spatial-temporal texture coefficient K^SPt_ext.

29. A method comprising: estimating visual quality parameters that indicate the visual quality of decompressed video frames; and substantially improving the visual quality of said decompressed video frames using their associated parameters.

30. The method of claim 29 wherein said visual quality is defined at least as the extent of distortions and ringing effects.

31. The method of claim 29, wherein said decompressed video frames are decompressed according to a decoding technique including at least one of: a joint photographic expert group (JPEG), a moving picture expert group (MPEG), a H- compression, and a Windows Media (WM).

32. The method of claim 29, wherein said estimating comprises: estimating a distortion level DR in a frame, an image complexity value NC, a per pixel intensity change hy and a high contrast small details level CH in a frame.

33. The method of claim 32, wherein said estimating comprises determining a blockness level.

34. The method of claim 33, wherein said determining comprises: summing high frequency components in every k pixels along a line of said frame; and detecting line-to-line periodicity by summing the output of said high frequency summing every k^Λ line, starting at different lines and by summing those blocks which are distorted.

35. The method of claim 32, wherein said estimating comprises: generating a per-pixel intensity change hy; and determining a per-frame image complexity level TSfC.

36. The method according to claim 35 and wherein said generating comprises: subtracting said decompressed video frame and a previous decompressed video frame to generate a difference frame ΔF; taking an absolute value of said difference frame Δ_F; and horizontally low pass filtering to reduce noise in the output of said absolute value unit.

37. The method according to claim 35 and wherein said determining comprises: subtracting a histogram of the intensities of said frame from that of a previous frame; generating an overall change value from said per-pixel intensity changes hy, said overall change value indicating the extent of significant change in said frame; and weighted summing the outputs of said subtracting and generating to generate said per-frame image complexity level NC.

38. The method of claim 32, wherein said estimating comprises: shifting said frequency components of said decompressed video frame downward by a first threshold, wherein said first threshold THDl is defined by expected intensity levels of high contrast details; limiting high frequency components of said decompressed video frame to an intensity level below a second threshold; two-dimensional low pass filtering the output of said limiting to detect dots in said frame; and limiting the output of said two-dimensional low pass filter to an intensity level below a third threshold.

39. The method of claim 32 wherein said substantially improving comprises: temporally adapting processing of each said decompressed video frame with at least one of said associated parameters; and spatially adapting processing of the output of said temporal adapting with at least one of said associated parameters.

40. The method of claim 39 and further comprising generating at least temporal control parameters and spatial-temporal control parameters from at least one of said visual quality parameters.

41. The method of claim 40, wherein said temporal control parameters comprise at least one of: a temporal threshold THD_TP determined by: THD _τp= 2[(l + DR^* )+ (l + NC)\ ; a temporal texture coefficient

= 1-DR^* ; a temporal contrast coefficient K^m determined by: K^_nt = 1-DR^* ; and a low frequency coefficient Kp determined by: K_F= 1 - 0.5Di?^* , wherein said DR^* is a normalized value of said distortion level DR.

42. The method of claim 40, wherein said spatial-temporal control parameters comprise at least one of: a spatial-temporal threshold THD_SP determined by: THD_SP = 3(1+Di?^*) , a per-pixel, recursion coefficient K_reo(i,j) determined by: KJU j) = K_^0 [l + 0.25(DR^* + CH,/)]; a per-pixel ringing coefficient K^ij) determined by: K_ring (z, J) = 1 - 0.5h_i} ^* , a spatial-temporal texture coefficient K^SP _text determined by: K?:_t = K_texto(l-DR^*); and a per-pixel, spatial-temporal contrast coefficient K^SF _oont(ij)

KL³ Q, j) = κ_cmt0 [i - o.5(/jR^* - cH,/)], wherein said DR is a normalized value of said distortion level DR, said CHy- are normalized values of said CH_y values and said h/ are normalized values of said h_y, and where said Kc₀nt,o is a maximum contrast coefficient, said Kreco is a maximum recursion coefficient, and said Ktexto is a maximum texture coefficient.

43. The method of claim 42, and comprising using said per-pixel, ringing coefficient Kring(i j) to reduce ringing effects in said decompressed frame.

44. The method of claim 41, wherein said temporally adapting comprises: temporally processing said decompressed video frame; first adjusting the sharpness of textual elements in said frame; second adjusting the sharpness of small details in said frame; and reducing blockness in said frame by attenuating low frequency components of said frame received from said temporal processing.

45. The method of claim 44 and wherein said first adjusting comprises: first limiting an input signal level below said temporal threshold THD_TP, wherein the input signal includes vertical high frequency components of said decompressed frame; second limiting an input signal level below said temporal threshold THD_TP, wherein the input signal includes horizontal high frequency components of said decompressed frame; horizontal low pass filtering of a signal produced by said first limiting; vertical low pass filtering of a signal produced by said second limiting; and adjusting texture sharpness of the sum of the outputs of said filtering steps according to said temporal texture coefficient

46. The method of claim 45 and wherein said adjusting comprises: generating a small details signal from said vertical and horizontal high frequency components and the outputs of said steps of limiting; and adjusting a sharpness of said small details signal according to said temporal contrast coefficient K TP cont-

47. The method of claim 45 and wherein said reducing blockness comprises reducing the level of low frequency components of said decompressed video frame according to said blockness coefficient K_F.

48. The method of claim 42, wherein said spatially adapting comprises: performing spatial processing on said decompressed video frame; adjusting textual elements in said frame; sharpening small details in said frame; and reducing ringing effects in said frame.

49. The method of claim 48, wherein said reducing comprises: first limiting an input signal level below said spatial threshold THD_SP, wherein the input signal includes vertical high frequency components of said decompressed frame; second limiting an input signal level below said spatial threshold THD_SP, wherein the input signal includes horizontal high frequency components of said decompressed frame; and reducing ringing effects in high pass filtered signals according to said ringing coefficient KώigGj).

50. The method of claim 48 wherein said sharpening comprises adjusting a sharpness of said small details signal according to said spatial-temporal contrast coefficient κ^: SP cont(Lj)-

51. The method of claim 48 and wherein said adjusting comprises: a temporally low pass filtering to reduce noise according to said recursion coefficient K_rec; and adjusting texture sharpness of an output of said filtering according to a value of said spatial-temporal texture coefficient K^SP _text.

52. The method of claim 39, wherein said temporally adapting comprises an infinite impulse response (IIR) filtering using an adaptive recursion coefficient determined at least from said estimated distortion level DR.

53. A computer product readable by a machine, tangibly embodying a program of instructions executable by the machine to implement a process, said process comprising: estimating visual quality parameters that indicate the visual quality of decompressed video frames; and substantially improving the visual quality of said decompressed video frames using their associated parameters.

54. The computer software product of claim 53, wherein said estimating comprises: estimating a distortion level DR in a frame, an image complexity value NC, a per pixel intensity change h^ and a high contrast small details level CH in a frame.

55. The computer software product of claim 54, wherein said estimating comprises determining a blockness level.

56. The computer software product of claim 54, wherein said estimating comprises: generating a per-pixel intensity change hy; and determining a per-frame image complexity level NC.

57. The computer software product of claim 54 wherein said substantially improving comprises: temporally adapting processing of each said decompressed video frame with at least one of said associated parameters; and spatially adapting processing of the output of said temporal adapting with at least one of said associated parameters.

58. The computer software product of claim 57, wherein said temporally adapting comprises: temporally processing said decompressed video frame; first adjusting the sharpness of textual elements in said frame; second adjusting the sharpness of small details in said frame; and reducing blockness in said frame by attenuating low frequency components of said frame received from said temporal processing.

59. The computer software product of claim 57, wherein said spatially adapting comprises: performing spatial processing on said decompressed video frame; adjusting textual elements in said frame; sharpening small details in said frame; and reducing ringing effects in said frame.

60. The computer product of claim 57, wherein said temporally adapting comprises an infinite impulse response (IIR) filtering using an adaptive recursion coefficient determined at least from said estimated distortion level DR.