WO2010058291A1 - Adaptive frequency domain filtering for phase plane correlation - Google Patents
Adaptive frequency domain filtering for phase plane correlation Download PDFInfo
- Publication number
- WO2010058291A1 WO2010058291A1 PCT/IB2009/007627 IB2009007627W WO2010058291A1 WO 2010058291 A1 WO2010058291 A1 WO 2010058291A1 IB 2009007627 W IB2009007627 W IB 2009007627W WO 2010058291 A1 WO2010058291 A1 WO 2010058291A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- phase plane
- high frequency
- window function
- images
- correlator
- Prior art date
Links
- 230000003044 adaptive effect Effects 0.000 title claims abstract description 35
- 238000001914 filtration Methods 0.000 title abstract description 26
- 238000000034 method Methods 0.000 claims abstract description 55
- 238000010606 normalization Methods 0.000 claims description 19
- 230000002596 correlated effect Effects 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 8
- 238000004519 manufacturing process Methods 0.000 claims description 4
- 238000013144 data compression Methods 0.000 claims description 3
- 238000004590 computer program Methods 0.000 claims 8
- 230000001131 transforming effect Effects 0.000 claims 6
- 230000033001 locomotion Effects 0.000 abstract description 29
- 239000013598 vector Substances 0.000 abstract description 20
- 230000008569 process Effects 0.000 abstract description 15
- 230000009286 beneficial effect Effects 0.000 abstract description 3
- 238000011045 prefiltration Methods 0.000 abstract description 3
- 230000006870 function Effects 0.000 description 57
- 238000010586 diagram Methods 0.000 description 11
- 230000000875 corresponding effect Effects 0.000 description 7
- 241000023320 Luma <angiosperm> Species 0.000 description 6
- 230000008901 benefit Effects 0.000 description 6
- 238000006243 chemical reaction Methods 0.000 description 6
- OSWPMRLSEDHDFF-UHFFFAOYSA-N methyl salicylate Chemical compound COC(=O)C1=CC=CC=C1O OSWPMRLSEDHDFF-UHFFFAOYSA-N 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 238000007906 compression Methods 0.000 description 4
- 230000006835 compression Effects 0.000 description 4
- 238000013459 approach Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 230000004069 differentiation Effects 0.000 description 3
- 238000006073 displacement reaction Methods 0.000 description 3
- 238000009826 distribution Methods 0.000 description 3
- 238000001228 spectrum Methods 0.000 description 3
- 230000002123 temporal effect Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000000750 progressive effect Effects 0.000 description 2
- 101100173586 Schizosaccharomyces pombe (strain 972 / ATCC 24843) fft2 gene Proteins 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000000873 masking effect Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000010363 phase shift Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000000135 prohibitive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
- H04N19/117—Filters, e.g. for pre-processing or post-processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/262—Analysis of motion using transform domain methods, e.g. Fourier domain methods
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N19/136—Incoming video signal characteristics or properties
- H04N19/14—Coding unit complexity, e.g. amount of activity or edge presence estimation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
- H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
- H04N19/176—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/48—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using compressed domain processing techniques other than decoding, e.g. modification of transform coefficients, variable length coding [VLC] data or run-length data
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
- H04N19/51—Motion estimation or motion compensation
- H04N19/547—Motion estimation performed in a transform domain
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/60—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/60—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
- H04N19/61—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding in combination with predictive coding
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/80—Details of filtering operations specially adapted for video compression, e.g. for pixel interpolation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
Definitions
- the invention relates in general to correlating two images. More specifically, the invention relates to the use of phase plane correlation (PPC).
- PPC phase plane correlation
- Motion vector data provides information related to the speed and direction of movements of critical parts of an image, e.g., portions of the image determined to be changing over a period of time such as from one frame of an image to the next.
- Applications making use of motion vector data include, but are not limited to, format conversion, de-interlacing, compression, image registration and others where some kind of temporal interpolation is necessary. Format conversion examples include 1) frame rate conversion, such as the conversion of NTSC video rate to HDTV video rate, 2) conversion of interlaced video to progressive video, and 3) the 3-to-2 pulldown artifact removal in conventional DVD format video.
- Video data compression processes also benefit from accurate motion vector data.
- Compression is often used to permit the useful transmission of data through a restricted bandwidth.
- Popular video compression algorithms utilize video compression standards, such as, for example, MPEG2, MPEG4, H.26L, etc.
- Another application that benefits from accurate motion vector data analysis is the production of display special effects, such as the global estimation of camera parameters useful to produce display effects for pan, tilt or zoom.
- the ITU-T International Telecommunication Union Tele-communication
- Standardization Sector recommends H.261 and H.262 standards as methods for encoding, storing, and transmitting image signals.
- the ISO International Organization for Standardization
- MPEG-I (11172-2)
- MPEG-2 13818-2
- Methods based on these standards adopt inter frame prediction for motion compensation in encoding video signals.
- Inter frame prediction is based upon the recognized redundancy characteristic of video data.
- Video signals contain highly redundant information from frame to frame (many image elements of a predetermined frame do not move and thus will be repeated in a subsequent frame). This holds true for frames generated as a result of special effects, for example, or frames generated to increase the definition of a video signal.
- Motion compensated inter frame prediction is a technique that takes advantage of inter frame redundancy to reduce the amount of data required to describe sequences of video frames or to create image frames, such as those created for example, in producing a progressive scan video signal from an interlaced video signal. An accurate determination of frame to frame motion is important to conduct such operations.
- One typical method for motion detection is carried out in the image domain. This method attempts to match blocks from a reference (previous) image frame with blocks from a current (subsequent to the reference) frame. Many so-called block matching methods start by calculating the absolute values of the differences in pixels in a block of a current image frame with all of the blocks in the reference image frame. A block in the current image frame having the smallest difference is determined to match. The displacement between the block in the current frame and the corresponding matching block in the reference frame is then characterized by horizontal and vertical displacement components, thus producing a motion vector. This procedure is known as the full-search procedure.
- Phase plane correlation is an efficient technique for correlating two images. In the frequency domain, motion is indicated by a phase shift (phase difference) between a particular block of a current image frame and a corresponding block of a reference image frame. A correlation surface obtained by an inverse Fourier transform of the phase difference indicates the quantity of pixels that moved and the magnitude of pixel movement.
- phase plane correlation motion vector determination techniques do not meet current video processing demands. There remains a need for an efficient method to calculate the motion in an image with a reduction in the chance for producing erroneous assignments of motion vectors to pixels.
- phase plane correlation results in a correlation surface in which peaks and their respective amplitudes are a direct indication of the similarity in image luminance.
- PPC phase plane correlation
- a standard luminance based PPC correlation surface is derived as follows:
- This equation is referred to as the "luminance based PPC”.
- I, F and F "1 are the image luminance, Fourier and inverse Fourier transforms, respectively.
- FIG. 1 is a block diagram illustrating a portion of a known arrangement for PPC image correlation.
- a luminance signal LUMA 1 which may represent a first image in the time domain, is transformed by a first Fast Fourier Transform FFTl into a first frequency domain signal Fl.
- a luminance signal LUMA 2 which may represent a second image, is transformed by a second Fast Fourier Transform FFT2 into a second frequency domain signal F2.
- Phase plane correlation is carried out in a defined window in a phase plane correlator 102 according to the expression
- phase plane correlator 102 (W.F1).(W.F2)* /
- a frequency domain signal from phase plane correlator 102 is transformed back to the time domain by an Inverse Fast Fourier Transform IFFT to provide a phase plane correlated surface.
- the invention is directed in general to improved Phase Plane Correlation (PPC) techniques and includes both methods and apparatus.
- PPC Phase Plane Correlation
- a PPC technique according to an embodiment of the invention uses adaptive frequency domain filtering to aid in the generation of candidate motion vectors when a PPC is used to determine image correlation.
- pre-filtering it is first determined whether or not it would be beneficial to pre-filter an input image prior to applying a conventional PPC process. If pre-filtering is determined to be desirable, it is carried out. If, on the other hand, pre-filtering is determined not to be needed, it is not carried out and the normal PPC process proceeds without any pre-filtering.
- the use of selective pre-filtering results in the creation of more reliable and consistent PPC surfaces than otherwise would be obtained without such pre-filtering.
- the decision regarding pre- filtering and the pre- filtering are carried out using an adaptive frequency domain filter. Pre-filtering is carried out in the frequency domain where time-domain convolution is effectively carried out efficiently using components that carry out multiplication with an in-place window. Energy of image high-frequency content is measured in computed Fourier surfaces to gauge the degree of high frequency content in the image. The amount of high frequency content is then used to control or "adapt" the filtering process.
- the conventional phase plane approach uses a fully normalized numerator. This is not always necessary depending on the application.
- normalizing the denominator used to "whiten" the frequency spectrum can be modulated to potentially deal with frequency band specific content more appropriately.
- the high-frequency estimator 316 one can choose to whiten to a lesser or greater degree based on the relative distribution of content in the frequency surface. For example, if the image content is largely high-frequencies, the exponent can be made to be larger than it would otherwise be. In contrast, if the image content is more narrow banded in nature (less high frequency components), one could choose a normalizing exponent that is smaller than it would otherwise be.
- a method includes: i) computing the Fourier transform of first and second images; ii) estimating a high-frequency content directly from the Fourier surfaces; iii) computing a window function that is a function of the estimated high- frequency energy; iv) applying the window to the Fourier surfaces; and v) providing modified Fourier surfaces into a PPC process.
- FIG. 1 is a block diagram illustrating a portion of a known arrangement for PPC image correlation.
- FIG. 2 is a flow chart of a method according to an embodiment of the invention.
- FIG. 3 is a block diagram of an arrangement for PPC image correlation including adaptive frequency domain filtering according to an embodiment of the invention.
- FIG. 4 is a graphical representation of an image in the frequency plane (either Fl or F2 in FIG. 2) illustrating the concept of high frequency estimation that occurs in high frequency estimator 314.
- FIG. 5 is a graphical representation of a result of high frequency estimation carried out by high frequency estimator 314.
- FIG. 6 is a block diagram of high frequency estimator 314 shown as a general block in FIG. 3.
- FIG. 7 is a block diagram of an embodiment of high frequency estimator 314 and adaptive window function generator 316.
- FIG. 3 is a block diagram of an embodiment of high frequency estimator 314 and adaptive window function generator 316.
- phase plane forming module 326 shown as a general block in FIG. 3.
- the invention provides a Phase Plane Correlation (PPC) technique that uses adaptive frequency domain filtering to aid in the generation of candidate motion vectors when the Phase Plane Correlation approach is used. It is determined whether or not it would be beneficial to pre-filter an input image prior to applying the PPC process. If so, pre-filtering is carried out using an adaptive frequency domain filter. This results in more reliable and consistent PPC surfaces than otherwise.
- PPC Phase Plane Correlation
- FIG. 2 is a flow chart of a method 200 of phase plane correlating according to an embodiment of the invention. It explains a general principle of the invention, namely optionally pre-filtering phase plane images to improve the correlation and accuracy of a phase plane correlated surface.
- a first image, represented by LUMA 1 in the time domain is transformed at step 210 into the frequency domain to form a frequency domain signal Fl. This step can be carried out by a Fast Fourier Transform.
- a second image, represented by LUMA 2 in the time domain is transformed at step 212 into the frequency domain to form a frequency domain signal F2. This step can also be carried out by a Fast Fourier Transform.
- An amount of high frequency energy in an image such as phase plane images Fl and F2 is determined at step 213.
- pre-filtering is determined based on the energy of high frequency components determined at 213. If pre-filtering is determined to be desirable, it is carried out at 216 using an adaptive filtering process. Adaptive filtering is carried out by generating and using an adaptive window that is generated based on an amount of high frequency content present in phase plane images Fl and F2. The amount of high frequency content present in phase plane images Fl and F2 is used to determine the shape of a window function used for pre- filtering. Note that in various embodiments of the present invention the adaptive window can be based on the high frequency content present in only phase plane image Fl, on the high frequency content present in only phase plane image F2, or based on the high frequency content present in both phase plane images Fl and F2.
- An adaptive window function can be controlled so that it is possible to choose to whiten to a lesser or greater degree based on a relative distribution of content in the frequency surfaces.
- High frequency content is preferably measured in an edge portion of phase plane images Fl and F2.
- the exponent can be made to be larger than it otherwise would be.
- the normalizing exponent can be made to be smaller than it would otherwise be.
- any profile or schedule of normalizing exponent could be devised depending on the exact nature and purpose that would make it possible to gauge the precision of, and the extent of differentiation, of candidate motion vectors.
- correlation is carried out at 220.
- the correlation result of 220 is inverse Fourier transformed at 222 to provide a phase plane correlated surface in the time domain at 224.
- FIG. 3 is a block diagram of an arrangement for PPC image correlation including adaptive frequency domain filtering according to an embodiment of the invention.
- a first image represented by a signal LUMA 1 in the time domain is transformed to the frequency domain by a Fast Fourier Transform module 310 to provide a phase plane image Fl.
- a second image, represented by a signal LUMA 2 in the time domain is transformed to the frequency domain by a Fast Fourier Transform module 312 to provide a phase plane image F2.
- Phase plane images Fl and F2 are input to a high frequency estimator 314 which is configured to estimate a total energy of a FFT surface and a high frequency energy of that FFT surface.
- High frequency estimator 314 thus provides both a total energy and a high frequency energy content of FFT surfaces Fl and F2.
- the total energy estimate is provided at an output 340 and the high frequency estimate is provided at an output 342.
- the process carried out by high frequency estimator 314 is explained further with respect to FIG. 6, described below.
- the total energy estimate and high frequency energy estimate are provided to adaptive window function generator 316.
- Adaptive window function generator 316 provides an adaptive window that is used to carry out adaptive filtering of phase plane images Fl and F2.
- This adaptive window is generated based on an amount of high frequency content present in phase plane images Fl and F2, as provided by high frequency estimator 314.
- the amount of high frequency content present in phase plane images Fl and F2 is used to determine the shape of a window function generated by adaptive window function generator 316.
- adaptive window function generator 316 computes a window function based on total and high frequency energy values and then generates a window based on the computed window function.
- Operation of high-frequency estimator 314 and adaptive window function generator 316 can be controlled so that it is possible to choose to whiten to a lesser or greater degree based on a relative distribution of content in the frequency surfaces.
- the exponent in the denominator term of Equation 2 (below) can be made to be larger than it otherwise would be.
- the normalizing exponent can be made to be smaller than it would otherwise be.
- any profile or schedule of normalizing exponent can be used depending on the exact nature and purpose of gauging the precision of, and the extent of differentiation, of candidate motion vectors.
- Window function generator 316 is configured to receive from high frequency estimator 314 estimates of total energy and high frequency energy in a FFT surface. These estimates are on lines 340 and 342, respectively, shown in FIG. 3. The total energy and high frequency energy are used to determine a weighting function that will be applied to a FFT surface.
- the exponent alpha is also determined from these two input values using a user-defined function in block 360.
- the user-defined function may be expressed by a closed-form analytic equation or it may be represented by a lookup table. Note that in various embodiments of the present invention the exponent alpha can be based on the high frequency content present in only phase plane image Fl, on the high frequency content present in only phase plane image F2, or based on the high frequency content present in both phase plane images Fl and F2.
- the high frequency content is preferably determined in an edge portion of phase plane images Fl and F2.
- the window function is represented by Wl and W2 at outputs of adaptive window function generator 316.
- Phase plane image Fl is windowed by the window function Wl by providing Fl and the window function at inputs of a mixer 318 which provides a first windowed surface at an output 320.
- Image F2 is windowed by the window function W2 by providing F2 and the window function at inputs of a mixer 322 to provide a second windowed surface at an output 324.
- Wl and W2 can be different functions or the same function denoted by the symbol W.
- phase plane forming module 326 processes the first and second windowed surfaces to form a phase plane based on the signals at an output 328 thereof.
- a more detailed diagram of phase plane forming module 326 is shown in FIG. 8. In essence, phase plane forming module 326 provides a frequency whitened spectrum. Operation of phase plane forming module 326 can be expressed by (W1.F1).(W2.F2)* /
- Output 328 of phase plane forming module 326 is coupled to an input of an
- Inverse Fast Fourier Transform module 330 which provides an image surface that has been phase plane correlated. This surface can then be fed to a conventional phase plane correlation (PPC) processing system.
- PPC phase plane correlation
- the adaptive window function provided by the adaptive window function generator 316 can be computed in various ways. An embodiment explaining how the window function may be computed is now set forth. According to this embodiment, the shape of the window function generated is based at least in part on the frequency content of the frequency domain surface as computed by the two-dimensional Fourier transforms at FFT 310 and FFT 312. According to an exemplary embodiment, one possible behavior for the window function would be to alter window shape based on the amount of high- frequency content present in the frequency domain surfaces. For example, it is possible to define a region of the frequency domain surface that corresponds to significant presence of high-frequency content and compute a ratio of this energy in that region to the total energy of the surface.
- Equation 1 represents a two-dimensional Gaussian window function.
- Equation 2 represents fractionally normalized phase plane correlation.
- Equation 2 includes an exponent on the denominator.
- the exponent on the denominator allows one to modulate the degree to which the spectrum in the frequency domain is "whitened.” For example, when the exponent is unity, the result is a conventional PPC. When the exponent is zero, the result is a conventional cross- correlation. However, when the exponent is between zero and one, a fractional normalization takes place.
- the exponent can also be greater than one, although such an exponent value does not correspond to a readily understood intuitive understanding.
- the degree to which normalization occurs depends on whether there is present narrow band or wide band noise. In general, the exponent can be any arbitrary function of the frequency surface.
- FIG. 4 is a graphical representation of a phase plane image 400, such as, for example phase plane image Fl or phase plane image F2.
- This figure helps to explain the concept of high frequency estimation that is carried out by high frequency estimator 314.
- Edge regions 410 of phase plane surface 400 represent high frequency regions of a Fast Fourier Transform (FFT) surface.
- FFT Fast Fourier Transform
- a motion vector would, for example be somewhere in a central region 412 of phase plane surface 400.
- the graphical representation is that of a matrix of frequency domain points. The energy of each such point located in an edge region, such as 410, is calculated. Then, the energies of such points are summed to determine a total energy present in "high frequency" components of the FFT surface. A signal representing this total energy is what is coupled from high frequency estimator 314 to an input of adaptive window function generator 316 in FIG. 3.
- FIG. 5 is a graphical representation of a result of high frequency estimation carried out by high frequency estimator 314.
- the curves represent examples of the shape of a window function generated by adaptive window function generator 316.
- the left portion of the FIG. 5 representation illustrates an exemplary window function for a case in which there is a low amount of energy in high frequency components of an image such as phase plane image Fl or phase plane image F2.
- the right portion of the FIG. 5 representation illustrates an exemplary window function for a case in which there is a higher amount of energy in high frequency components of a phase plane image such as phase plane image Fl or phase plane image F2.
- FIG. 6 is a block diagram further explaining operation of high frequency estimator
- a frequency surface F such as Fl produced by FFT 310 or F2 produced by FFT 312 is parsed by a parsing module 602 and a parsing module 604.
- a high frequency computation module 606 computes an estimate of high- frequency energy using a user-defined mask, or weighting function. This high frequency energy estimate is provided at output 342 of high frequency estimator 314.
- a total energy computation module 608 computes an estimate of total energy using a user-defined mask, or weighting function. This total energy estimate is provided at output 340 of high frequency estimator 314.
- the function can, for example, be a simple weighting with real- valued numbers corresponding to elements in the FFT surface.
- the energy calculation can be, for example, the magnitude of a complex phasor at each entry of the FFT array. Parsing and computation may be performed serially or concurrently. These functions can be carried out by software running on a processor or by a hard-wired circuit.
- FIG. 7 is a block diagram of an embodiment of high frequency estimator 314 and adaptive window function generator 316.
- An FFT surface is generated by a FFT, such as, for example FFT 310 shown in FIG. 3. This surface is represented by "S" in FIG. 7.
- S is a frequency domain surface that can be represented by a complex phasor S as shown in the figure.
- Surface S is input to high frequency estimator 314 which, in this embodiment, comprises total energy estimator 704 and high frequency estimator 706.
- Total energy estimator 704 provides an estimate El of total energy of surface S.
- High frequency estimator 706 provides an estimate E2 of high frequency energy of surface S, such as, for example, in region 410 shown in FIG. 4.
- El represents the measure of activity in the signal at each frequency.
- Adaptive window function generator 316 includes a window function generator 708 and a window generator 710.
- a window function is computed by window function generator 708 based on energy El and energy E2.
- the function generated by window function generator 708 is used by window generator 710 to generate an actual window (to be input to mixer 318 or 322 shown in FIG. 3).
- Equation 3 represents weighted phase plane correlation with fractional normalization.
- the exponent is computed by examining the ratio produced in FIG. 6 outputs 342 versus 340, or by means of a look-up table - as described previously. This forms the exponent term as shown in the equation above.
- FIG. 8 is a diagram further explaining operation of phase plane forming module
- Phase plane forming module 326 receives inputs 320 and 324 and provides an output 328.
- the internal components of phase plane forming module 326 comprised of the inputs 320 and 324 and output 328 consists of a Euclidean norm operation on the complex vector 800, the exponentiator 801 and inverter
- inverter 802 which collectively carry out the functions represented by the denominator of Equation 3, above, representing weighted phase plane correlation with fractional normalization.
- the output 805 of inverter 802 is multiplied by a multiplier 803 with a signal 806 from a multiplier 820 receiving inputs 320 and 324.
- simulation, synthesis and/or manufacture of the various embodiments of this invention may be accomplished, in part, through the use of computer readable code, including general programming languages (such as C or C++), hardware description languages (HDL) including Verilog HDL, VHDL, Altera HDL (AHDL) and so on, or other available programming and/or schematic capture tools (such as circuit capture tools).
- general programming languages such as C or C++
- HDL hardware description languages
- AHDL Altera HDL
- circuit capture tools such as circuit capture tools
- This computer readable code can be disposed in any known computer usable medium including semiconductor, magnetic disk, optical disk (such as CD-ROM, DVD-ROM) and as a computer data signal embodied in a computer usable (e.g., readable) transmission medium (such as a carrier wave or any other medium including digital, optical, or analog-based medium).
- a computer usable (e.g., readable) transmission medium such as a carrier wave or any other medium including digital, optical, or analog-based medium.
- the code can be transmitted over communication networks including the Internet and internets.
- a core such as a GPU core
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Image Analysis (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
Description
Claims
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP09827244A EP2359335A4 (en) | 2008-11-20 | 2009-11-20 | Adaptive frequency domain filtering for phase plane correlation |
JP2011536967A JP5603341B2 (en) | 2008-11-20 | 2009-11-20 | Adaptive frequency domain filtering for phase-plane correlation |
CN200980154534.7A CN102282848B (en) | 2008-11-20 | 2009-11-20 | Adaptive frequency domain filtering for phase plane correlation |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US11649308P | 2008-11-20 | 2008-11-20 | |
US61/116,493 | 2008-11-20 | ||
US12/621,924 | 2009-11-19 | ||
US12/621,924 US9300966B2 (en) | 2008-11-20 | 2009-11-19 | Adaptive frequency domain filtering for phase plane correlation |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2010058291A1 true WO2010058291A1 (en) | 2010-05-27 |
Family
ID=42197875
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/IB2009/007627 WO2010058291A1 (en) | 2008-11-20 | 2009-11-20 | Adaptive frequency domain filtering for phase plane correlation |
Country Status (6)
Country | Link |
---|---|
US (1) | US9300966B2 (en) |
EP (1) | EP2359335A4 (en) |
JP (1) | JP5603341B2 (en) |
KR (1) | KR20110097868A (en) |
CN (1) | CN102282848B (en) |
WO (1) | WO2010058291A1 (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104917931B (en) * | 2015-05-28 | 2018-03-02 | 京东方科技集团股份有限公司 | Moving image compensation method and device, display device |
US9892518B2 (en) * | 2015-06-09 | 2018-02-13 | The Trustees Of Columbia University In The City Of New York | Systems and methods for detecting motion using local phase information |
WO2017100666A1 (en) * | 2015-12-09 | 2017-06-15 | Cohere Technologies | Pilot packing using complex orthogonal functions |
WO2017147439A1 (en) | 2016-02-25 | 2017-08-31 | Cohere Technologies | Reference signal packing for wireless communications |
GB2574428B (en) * | 2018-06-06 | 2020-06-10 | Visidon Oy | Image registering method and apparatus |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040062420A1 (en) * | 2002-09-16 | 2004-04-01 | Janos Rohaly | Method of multi-resolution adaptive correlation processing |
US20040179594A1 (en) * | 2003-02-20 | 2004-09-16 | The Regents Of The University Of California | Phase plane correlation motion vector determination method |
Family Cites Families (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
GB9325922D0 (en) * | 1993-12-18 | 1994-02-23 | Kodak Ltd | Detection of global translations between images |
US6401057B1 (en) * | 1997-07-02 | 2002-06-04 | Nonlinear Solutions, Inc. | Detection and classification system for analyzing deterministic properties of data using correlation parameters |
EP1050850A1 (en) * | 1999-05-03 | 2000-11-08 | THOMSON multimedia | Process for estimating a dominant motion between two frames |
US6230106B1 (en) * | 1999-10-13 | 2001-05-08 | Modulation Instruments | Method of characterizing a device under test |
US6947509B1 (en) * | 1999-11-30 | 2005-09-20 | Verance Corporation | Oversampled filter bank for subband processing |
US20050033505A1 (en) * | 2002-12-05 | 2005-02-10 | Premier Wireless, Inc. | Traffic surveillance and report system |
US7949148B2 (en) * | 2006-01-23 | 2011-05-24 | Digimarc Corporation | Object processing employing movement |
US7653230B2 (en) * | 2006-02-21 | 2010-01-26 | General Electric Company | Methods and systems for image reconstruction using low noise kernel |
JP4821548B2 (en) * | 2006-10-02 | 2011-11-24 | コニカミノルタホールディングス株式会社 | Image processing apparatus, image processing apparatus control method, and image processing apparatus control program |
KR101156117B1 (en) * | 2007-08-02 | 2012-07-03 | 삼성전자주식회사 | Apparatus and method for detecting video |
-
2009
- 2009-11-19 US US12/621,924 patent/US9300966B2/en active Active
- 2009-11-20 KR KR1020117014242A patent/KR20110097868A/en not_active Application Discontinuation
- 2009-11-20 WO PCT/IB2009/007627 patent/WO2010058291A1/en active Application Filing
- 2009-11-20 JP JP2011536967A patent/JP5603341B2/en active Active
- 2009-11-20 EP EP09827244A patent/EP2359335A4/en not_active Ceased
- 2009-11-20 CN CN200980154534.7A patent/CN102282848B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040062420A1 (en) * | 2002-09-16 | 2004-04-01 | Janos Rohaly | Method of multi-resolution adaptive correlation processing |
US20040179594A1 (en) * | 2003-02-20 | 2004-09-16 | The Regents Of The University Of California | Phase plane correlation motion vector determination method |
Non-Patent Citations (1)
Title |
---|
See also references of EP2359335A4 * |
Also Published As
Publication number | Publication date |
---|---|
US20100150225A1 (en) | 2010-06-17 |
JP5603341B2 (en) | 2014-10-08 |
CN102282848A (en) | 2011-12-14 |
KR20110097868A (en) | 2011-08-31 |
JP2012509535A (en) | 2012-04-19 |
CN102282848B (en) | 2015-04-29 |
EP2359335A1 (en) | 2011-08-24 |
EP2359335A4 (en) | 2012-08-01 |
US9300966B2 (en) | 2016-03-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP2074586B1 (en) | Image enhancement | |
US7664301B2 (en) | Method and apparatus for enhancing image quality of a two-dimensional ultrasound image | |
US6128047A (en) | Motion estimation process and system using sparse search block-matching and integral projection | |
Moorthy et al. | Efficient motion weighted spatio-temporal video SSIM index | |
JPH01500238A (en) | Method and device for measuring motion in television images | |
EP2214137B1 (en) | A method and apparatus for frame interpolation | |
US20030086498A1 (en) | Apparatus and method of converting frame and/or field rate using adaptive motion compensation | |
US20030081144A1 (en) | Video data de-interlacing using perceptually-tuned interpolation scheme | |
US7197074B2 (en) | Phase plane correlation motion vector determination method | |
US9280828B2 (en) | Image processing apparatus, image processing method, and program | |
US8243194B2 (en) | Method and apparatus for frame interpolation | |
WO2000062556A9 (en) | Method and apparatus for estimating digital video quality without using a reference video | |
US9300966B2 (en) | Adaptive frequency domain filtering for phase plane correlation | |
JPH0993614A (en) | Three-dimension image coder | |
Zeng et al. | 3D-SSIM for video quality assessment | |
US20140028911A1 (en) | Interpolation of images | |
Dane et al. | Optimal temporal interpolation filter for motion-compensated frame rate up conversion | |
US20060159177A1 (en) | Motion estimation method, device, and system for image processing | |
US20130301928A1 (en) | Shift vector reliability determining apparatus and method | |
Biswas et al. | A novel motion estimation algorithm using phase plane correlation for frame rate conversion | |
EP1222614B1 (en) | Method and apparatus for adaptive class tap selection according to multiple classification | |
Okarma | Colour image quality assessment using the combined full-reference metric | |
CN103618904B (en) | Motion estimation method and device based on pixels | |
EP2922294B1 (en) | System and methods to efficiently measure stereoscopic disparity and to generate related mismatch images | |
Takahashi et al. | An arbitrary scale image enlargement method with the prediction of high‐frequency components |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 200980154534.7 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 09827244 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2011536967 Country of ref document: JP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 3520/CHENP/2011 Country of ref document: IN |
|
ENP | Entry into the national phase |
Ref document number: 20117014242 Country of ref document: KR Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2009827244 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |