WO2012147247A1 - Dispositif d'affichage vidéo, procédé d'affichage vidéo et dispositif de traitement vidéo - Google Patents

Dispositif d'affichage vidéo, procédé d'affichage vidéo et dispositif de traitement vidéo Download PDF

Info

Publication number
WO2012147247A1
WO2012147247A1 PCT/JP2012/000786 JP2012000786W WO2012147247A1 WO 2012147247 A1 WO2012147247 A1 WO 2012147247A1 JP 2012000786 W JP2012000786 W JP 2012000786W WO 2012147247 A1 WO2012147247 A1 WO 2012147247A1
Authority
WO
WIPO (PCT)
Prior art keywords
block
motion vector
pixel
depth information
unit
Prior art date
Application number
PCT/JP2012/000786
Other languages
English (en)
Japanese (ja)
Inventor
森田　友子
Original Assignee
パナソニック株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by パナソニック株式会社 filed Critical パナソニック株式会社
Publication of WO2012147247A1 publication Critical patent/WO2012147247A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/50—Depth or shape recovery
    • G06T7/529—Depth or shape recovery from texture
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/10—Processing, recording or transmission of stereoscopic or multi-view image signals
    • H04N13/106—Processing image signals
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N2013/0074—Stereoscopic image analysis
    • H04N2013/0081—Depth or disparity estimation from stereoscopic image signals
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N2013/0074—Stereoscopic image analysis
    • H04N2013/0085—Motion estimation from stereoscopic image signals

Definitions

  • the present invention detects a motion vector from a plurality of frame images in a video signal consisting of a plurality of temporally continuous frame images, and displays the video signal by performing predetermined image processing using the detected motion vectors.
  • the present invention relates to a video display device, a video display method, and a video processing device.
  • LCDs liquid crystal displays
  • PDPs plasma displays
  • block matching processing is known as a method for detecting a motion vector (see, for example, Patent Document 1).
  • one frame image is divided into a plurality of blocks in units of 4 ⁇ 4 pixels, for example, and a motion vector is detected for each block.
  • FIG. 25 is a diagram for explaining detection of a conventional motion vector.
  • a soccer ball image 203 as a foreground image passes in front of a spectator seat image 201 and a lawn image 202 as background images, and the soccer ball image 203 is displayed at the center of the display screen 200. Has been.
  • the frame image is divided into a plurality of blocks, so that the display screen 200 is divided into, for example, 25 blocks as shown in FIG.
  • the motion vector detection range 204 corresponding to the soccer ball image 203 is nine blocks in the center of the screen.
  • the motion vector detection range 204 is larger than the soccer ball image 203, and an incorrect motion vector is detected in the area between the outer edge of the soccer ball image 203 and the outer edge of the detection range 204. Become.
  • FIG. 26 is a diagram for describing moving image noise that occurs when image processing is performed using an incorrect motion vector.
  • a car image 302 that is a foreground image passes in front of a building image 301 that is a background image.
  • a hatched area in FIG. 26 is a motion vector detection range 303 corresponding to the car image 302. As shown in FIG. 26, the motion vector detection range 303 is larger than the vehicle image 302.
  • an incorrect motion vector is detected in a region between the outer edge of the vehicle image 302 and the outer edge of the detection range 303.
  • moving image noise called “Halo” occurs, and the building image 301 that is the background image is distorted. It is displayed.
  • the present invention has been made to solve the above-described problem, and is capable of detecting a motion vector with high accuracy and preventing image quality deterioration, a video display method, and a video processing device. Is intended to provide.
  • a video display device detects a motion vector from a plurality of frame images in a video signal including a plurality of temporally continuous frame images, and uses the detected motion vectors to generate the video signal.
  • a video display device that performs predetermined image processing and displays, a motion vector detection unit that detects a motion vector of a current frame image using at least two or more frame images that are temporally changed, and the current
  • a depth information acquisition unit that acquires depth information representing a position in the depth direction of each pixel constituting the frame image, and the motion vector detection unit that is detected based on the depth information acquired by the depth information acquisition unit
  • a motion vector correction unit that corrects the motion vector.
  • the motion vector detection unit detects the motion vector of the current frame image using at least two or more frame images that are temporally changed.
  • the depth information acquisition unit acquires depth information indicating the position in the depth direction of each pixel constituting the current frame image. Then, the detected motion vector is corrected by the motion vector correction unit based on the acquired depth information.
  • the foreground image and the background image are clearly distinguished by the depth information representing the position in the depth direction, and the motion vector of the pixel at the boundary portion between the foreground image and the background image is corrected. It is possible to detect a motion vector and to prevent image quality deterioration.
  • (A) is a figure which shows an example of the high pass filter which has a characteristic (weighting coefficient) in a horizontal direction
  • (B) is a figure which shows an example of the high pass filter which has a characteristic (weighting coefficient) in a vertical direction
  • (C) is a diagram illustrating an example of a high-pass filter having characteristics (weighting coefficients) in the horizontal and vertical directions. It is a figure for demonstrating the additional process which sets a 1st threshold value based on the frequency of each pixel. It is a figure which shows the structure of the video display apparatus in the 2nd Embodiment of this invention. It is a figure which shows the structure of the video display apparatus in the 3rd Embodiment of this invention.
  • a liquid crystal display device will be described as an example of a video display device.
  • the video display device to which the present invention is applied is not particularly limited to this example, and examples thereof include a plasma display and an organic EL display. The same applies to the above.
  • FIG. 1 is a diagram showing a configuration of a video display apparatus according to the first embodiment of the present invention.
  • 1 includes a video signal input unit 1, a depth information acquisition unit 2, a motion vector detection unit 3, a motion vector correction unit 4, a video signal processing unit 5, and a display unit 6.
  • the video processing device includes a video signal input unit 1, a depth information acquisition unit 2, a motion vector detection unit 3, a motion vector correction unit 4, and a video signal processing unit 5.
  • the video signal input unit 1 includes, for example, a tuner for TV broadcasting, an image input terminal, a network connection terminal, and the like, and a video signal is input to the video signal input unit 1.
  • the video signal input unit 1 performs a known conversion process on the input video signal, and outputs the frame image data after the conversion process to the motion vector detection unit 3, the motion vector correction unit 4, and the video signal processing unit 5.
  • the depth information acquisition unit 2 acquires depth information indicating the position in the depth direction of each pixel constituting the current frame image.
  • the depth information is input from the outside to the video display device 10 together with the video signal as data accompanying the video signal.
  • the depth information acquisition unit 2 acquires depth information input from the outside and outputs the depth information to the motion vector correction unit 4.
  • the motion vector detection unit 3 detects the motion vector of the current frame image using at least two or more frame images that change in time.
  • the motion vector detection unit 3 is inputted with two temporally continuous frame image data, for example, image data of frame N-1 and image data of frame N (where N is an integer), and the motion vector detection unit 3 Detects a motion amount between these frames to detect a motion vector for each pixel of the frame N and outputs it to the motion vector correction unit 4.
  • this motion vector detection method a known motion vector detection method is used. For example, a detection method by block matching processing is used.
  • the motion vector correction unit 4 corrects the motion vector detected by the motion vector detection unit 3 based on the depth information acquired by the depth information acquisition unit 2. Details of the motion vector correction unit 4 will be described later.
  • the video signal processing unit 5 performs predetermined image processing on the video signal using the motion vector corrected by the motion vector correction unit 4.
  • the video signal processing unit 5 performs motion compensation based on the motion vector corrected by the motion vector correction unit 4.
  • the video signal processing unit 5 performs motion compensation processing based on the motion vector corrected by the motion vector correction unit 4, and generates interpolated frame image data that is interpolated between temporally preceding and following frames.
  • the generated interpolated frame image data is interpolated between frames.
  • the video signal processing unit 5 converts the frame rate (number of frames) by interpolating an image between frames, and improves motion blur.
  • the target frame image data and the previous frame image data are divided into macro blocks (for example, blocks of 16 pixels ⁇ 16 lines), and the target frame image data and the previous frame image data are divided.
  • Frame image data between frames is predicted from previous frame image data based on the motion vector indicating the moving direction and moving amount of the corresponding macroblock.
  • the video signal processing unit 5 generates interpolation frame image data to be interpolated between frames by motion compensation using the motion vector corrected by the motion vector correction unit 4, and the generated interpolation frame image data is By sequentially outputting together with the input image, the frame rate of the input image is converted from 60 frames / second to 120 frames / second, for example.
  • the display unit 6 displays the video signal that has been subjected to predetermined image processing by the video signal processing unit 5.
  • the display unit 6 includes a color filter, a polarizing plate, a backlight device, a liquid crystal panel, a panel drive circuit, and the like. Based on the frame image data interpolated by the video signal processing unit 5, a display signal and a data signal are displayed on the liquid crystal panel. When applied, a moving image is displayed.
  • FIG. 2 is a diagram showing a configuration of the motion vector correction unit shown in FIG.
  • the motion vector correction unit 4 illustrated in FIG. 2 includes a depth information difference value detection unit 41, a frequency detection unit 42, a first block matching difference value selection unit 43, a block matching threshold value setting unit 44, a first comparison unit 45, and a correction.
  • a target pixel selection unit 46, a block search unit 47, and a motion vector change unit 48 are provided.
  • the motion vector detection unit 3 divides the frame image into a plurality of divided blocks including a plurality of pixels, detects a motion vector for each divided block, and selects a motion vector before correction of each pixel as a first motion vector. Output to the unit 481 and the second motion vector selection unit 482. In addition, the motion vector detection unit 3 calculates a block matching difference value of each divided block and outputs the block matching difference value to the first block matching difference value selection unit 43 and the second block matching difference value selection unit 472.
  • the motion vector detection unit 3 calculates the absolute difference between the luminance value of each pixel of the target divided block of the current frame image and the luminance value of each pixel of the reference divided block referred to when detecting the motion vector of the past frame image. A value is calculated, and a total value of the calculated difference absolute values is calculated as a block matching difference value of the target divided block.
  • the motion vector detection unit 3 extracts a plurality of candidate vectors as motion vector candidates when detecting a motion vector.
  • the motion vector detection unit 3 calculates the absolute value of the difference between the luminance value of each pixel of the target divided block of the current frame image and the luminance value of each pixel of each reference divided block corresponding to a plurality of candidate vectors of the past frame image. And the difference absolute value is summed for each reference divided block.
  • the motion vector detection unit 3 detects, as a motion vector, a candidate vector corresponding to a reference divided block having a minimum total value among a plurality of candidate vectors. In addition, the motion vector detection unit 3 calculates the minimum total value as the block matching difference value of the target divided block.
  • the block matching difference value is an index of motion vector accuracy (reliability). If the block matching difference value is small, the motion vector accuracy is high, and if the block matching difference value is large, the motion vector accuracy is low. In the present embodiment, a divided block with low motion vector accuracy is selected based on the block matching difference value.
  • the motion vector detection unit 3 calculates block matching difference values for all the divided blocks of the current frame image.
  • the depth information difference value detection unit 41 calculates, for each pixel, the maximum difference absolute value between the depth information of the target pixel among the pixels constituting the current frame image and the depth information of a plurality of adjacent pixels adjacent to the target pixel. To calculate.
  • the frequency detector 42 detects the frequency of each pixel of the current frame image.
  • the first block matching difference value selection unit 43 selects a block matching difference value of the target divided block. Note that the first block matching difference value selection unit 43 selects a target divided block in order from the upper left divided block, for example, in the horizontal direction or the vertical direction of the frame image.
  • the block matching threshold value setting unit 44 sets a second threshold value Tc for comparison with the block matching difference value of the target divided block.
  • the second threshold value Tc is an area of 5% or more of the total area of one frame image, and the block matching difference value in an area composed of the target divided block and a plurality of divided blocks around the target divided block. For example, 1.2 times the average value. If the area is set to be narrow, pixels to be corrected are not detected when the accuracy of the motion vector of the entire area is low. Therefore, the area is sufficiently wide (for example, 5% of the total area of one frame image). % Or more) is set.
  • the block matching threshold value setting unit 44 sets the second threshold value Tc so that the value decreases as the maximum absolute value of the difference in depth information of each pixel calculated by the depth information difference value detection unit 41 increases. To do. Further, the block matching threshold value setting unit 44 sets the second threshold value Tc so that the value decreases as the frequency of each pixel detected by the frequency detection unit 42 increases.
  • the block matching difference value of the target divided block selected by the first block matching difference value selection unit 43 is equal to or greater than the second threshold value Tc set by the block matching threshold value setting unit 44. Determine whether or not.
  • the correction target pixel selection unit 46 determines whether or not the calculated maximum absolute value of the difference in the depth information of each pixel is equal to or greater than a predetermined first threshold Ta.
  • the first threshold Ta is preferably 32, for example.
  • the correction target pixel selection unit 46 has a maximum depth information difference absolute value calculated by the depth information difference value detection unit 41 that is equal to or greater than a predetermined first threshold Ta and is calculated by the motion vector detection unit 3. A pixel whose block matching difference value is equal to or greater than a predetermined second threshold Tc is selected as a correction target pixel for correcting the motion vector.
  • the correction target pixel selection unit 46 determines that the maximum absolute value of the difference in depth information is equal to or greater than the first threshold Ta, and the first comparison unit 45 determines that the block matching difference value is equal to or greater than the second threshold Tc.
  • a pixel included in the divided block determined to be present is selected as a correction target pixel.
  • the block search unit 47 increases the block matching difference value from the target correction target pixel among the correction target pixels selected by the correction target pixel selection unit 46 in a direction in which the maximum absolute value of the difference in depth information decreases.
  • a divided block that is equal to or less than a predetermined third threshold value Ts is searched for within each pixel to be corrected within a predetermined search range.
  • the block search unit 47 determines whether or not the maximum difference absolute value of the depth information of each pixel calculated by the depth information difference value detection unit 41 is equal to or greater than a predetermined fourth threshold value. When it is determined that the maximum difference absolute value of the depth information is greater than or equal to a predetermined fourth threshold, the search range is increased.
  • the block search unit 47 determines whether or not the frequency of each pixel detected by the frequency detection unit 42 is equal to or higher than a predetermined fifth threshold, and the frequency of each pixel is equal to or higher than a predetermined fifth threshold. If it is determined that there is, the search range is increased.
  • the block search unit 47 includes a search block identification unit 471, a second block matching difference value selection unit 472, a search block threshold value setting unit 473, and a second comparison unit 474.
  • the search block specifying unit 471 specifies search divided blocks that exist in the direction in which the maximum value of the absolute value of difference in depth information decreases from the target correction target pixels among the correction target pixels selected by the correction target pixel selection unit 46. To do.
  • the second block matching difference value selection unit 472 selects the block matching difference value of the search divided block specified by the search block specifying unit 471.
  • the search block threshold value setting unit 473 sets a third threshold value Ts for comparison with the block matching difference value of the search divided block.
  • the third threshold value Ts is smaller than the second threshold value Tc and is set so that the difference absolute value per pixel is 2, for example, and is 32 when one divided block is configured by 16 pixels.
  • the search block threshold value setting unit 473 determines whether or not the maximum value of the difference absolute value of the depth information of each pixel calculated by the depth information difference value detection unit 41 is equal to or greater than a predetermined fourth threshold value. If it is determined that the maximum difference absolute value of the depth information of each pixel is equal to or greater than a predetermined fourth threshold, the third threshold Ts is increased. Further, the search block threshold value setting unit 473 determines whether or not the frequency of each pixel detected by the frequency detection unit 42 is equal to or higher than a predetermined fifth threshold value, and the frequency of each pixel is set to a predetermined fifth threshold value. If it is determined that this is the case, the third threshold value Ts is increased.
  • the correction target pixel selection unit 46 determines whether or not the frequency of each pixel detected by the frequency detection unit 42 is equal to or higher than a predetermined fifth threshold, and the frequency of each pixel is a predetermined fifth threshold. If it is determined that the above is true, the first threshold Ta is decreased.
  • the block matching difference value of the search divided block selected by the second block matching difference value selection unit 472 is equal to or less than the third threshold value Ts set by the search block threshold value setting unit 473. Determine whether or not.
  • the motion vector change unit 48 converts the motion vector of the target correction target pixel into the motion vector of the divided block found by the block search unit 47. change.
  • the motion vector changing unit 48 does not change the motion vector of the target correction target pixel when the block search unit 47 does not find a divided block equal to or smaller than the third threshold Ts.
  • the motion vector change unit 48 includes a first motion vector selection unit 481, a second motion vector selection unit 482, a switch 483, and a selection determination unit 484.
  • the first motion vector selection unit 481 selects the motion vector of the current divided block and outputs it to the switch 483.
  • the second motion vector selection unit 482 selects the motion vector of the search divided block specified by the search block specification unit 471 and outputs it to the switch 483.
  • the switch 483 switches between the motion vector of the current divided block selected by the first motion vector selection unit 481 and the motion vector of the search divided block selected by the second motion vector selection unit 482.
  • the selection determination unit 484 sets the switch 483 to output the motion vector of the search division block.
  • the motion vector of the switching divided search block is output as the motion vector of the target correction target pixel.
  • the selection determination unit 484 determines the search division block when the second comparison unit 474 determines that the block matching difference value of the search division block is greater than the third threshold Ts, or the search block identification unit 471 identifies the search division block. If not, the switch 483 is switched to output the motion vector of the current divided block, and the motion vector of the current divided block is output as the motion vector of the target correction target pixel.
  • the selection determination unit 484 selects a motion vector of the current divided block or a motion vector of the search divided block for the correction target pixel among the pixels constituting the current frame image, and selects other than the correction target pixel. For the pixel, the motion vector of the current divided block is selected.
  • the motion vector detection unit 3 corresponds to an example of a motion vector detection unit
  • the depth information acquisition unit 2 corresponds to an example of a depth information acquisition unit
  • the motion vector correction unit 4 corresponds to a motion vector.
  • the depth information difference value detection unit 41 corresponds to an example of a depth information difference value calculation unit
  • the correction target pixel selection unit 46 corresponds to an example of a correction target pixel selection unit
  • the block search unit 47 corresponds to an example of a correction unit.
  • the motion vector change unit 48 corresponds to an example of a motion vector change unit
  • the search block threshold setting unit 473 corresponds to an example of a first threshold setting unit and a third threshold setting unit.
  • the frequency detection unit 42 corresponds to an example of a frequency detection unit
  • the block matching threshold setting unit 44 corresponds to an example of a second threshold setting unit and a fourth threshold setting unit.
  • FIG. 3 is a flowchart for explaining the operation of the video display apparatus according to the first embodiment of the present invention.
  • step S1 the video signal input unit 1 performs a known conversion process on the input video signal, and converts the frame image data after the conversion process into a motion vector detection unit 3, a motion vector correction unit 4, and a video signal process. Output to unit 5.
  • FIG. 4 is a diagram illustrating an example of a frame image according to the first embodiment of the present invention.
  • the foreground image SG is displayed on the background image BG, and the foreground image SG moves in the direction of the arrow in FIG.
  • the frame image FG is divided into 25 (5 ⁇ 5) divided blocks, and one divided block is configured by 16 (4 ⁇ 4) pixels.
  • processing for correcting a motion vector in the frame image FG illustrated in FIG. 4 will be described.
  • step S2 the depth information acquisition unit 2 acquires depth information representing the position in the depth direction of each pixel constituting the current frame image.
  • FIG. 5 is a diagram schematically illustrating depth information in the first embodiment of the present invention. As shown in FIG. 5, the depth information of the pixel corresponding to the foreground image SG is smaller than the depth information of the pixel corresponding to the background image BG.
  • step S3 the motion vector detection unit 3 detects the motion vector of the current frame image using at least two or more frame images that are temporally changed.
  • the motion vector detection unit 3 detects a motion vector for each divided block.
  • the motion vector detection unit 3 determines the luminance value of each pixel in the target divided block of the current frame image and the luminance value of each pixel of the reference divided block referred to when detecting the motion vector of the past frame image. Is calculated for each of the divided blocks as a block matching difference value of the target divided block. The motion vector detection unit 3 calculates block matching difference values of all the divided blocks of the current frame image.
  • FIG. 7 is a diagram schematically showing a block matching difference value in the first embodiment of the present invention.
  • the block matching difference value of a divided block around the frame image FG that is, a divided block including only the background image BG is 0, and only the divided block at the center of the frame image FG, that is, only the foreground image SG.
  • the block matching difference value of the included divided block is 0, and the block matching difference value of the divided block including the foreground image SG and the background image BG is 80.
  • step S4 the motion vector correction unit 4 performs a motion vector correction process for correcting the motion vector detected by the motion vector detection unit 3 based on the depth information acquired by the depth information acquisition unit 2.
  • FIG. 8 is a flowchart for explaining the motion vector correction processing in step S4 of FIG.
  • step S11 the depth information difference value detection unit 41 calculates the maximum difference absolute value between the depth information of the target pixel and the depth information of adjacent pixels adjacent to the target pixel.
  • FIG. 9 is a diagram schematically showing the maximum value of the absolute difference value of the depth information according to the first embodiment of the present invention. The maximum value of the difference absolute value of the depth information shown in FIG. 9 is calculated based on the depth information shown in FIG. The depth information difference value detection unit 41 calculates the maximum value of the absolute difference between the target pixel and the adjacent pixels adjacent to the target pixel in the vertical and horizontal directions.
  • step S12 the first comparison unit 45 determines whether or not the block matching difference value of the target divided block is equal to or greater than the second threshold Tc, thereby determining a block that is equal to or greater than the second threshold Tc.
  • a divided block having a matching difference value is extracted.
  • the block matching threshold value setting unit 44 has a second threshold value that is 1.2 times the average value of the block matching difference values in an area formed by the target divided block and a plurality of divided blocks around the target divided block Set Tc. As a result, the divided blocks indicated by hatching in FIG. 7 are extracted.
  • step S ⁇ b> 13 the correction target pixel selection unit 46 determines that the maximum depth information difference absolute value calculated by the depth information difference value detection unit 41 is greater than or equal to a predetermined first threshold Ta and a motion vector. A pixel whose block matching difference value calculated by the detection unit 3 is greater than or equal to a predetermined second threshold Tc is selected as a correction target pixel for correcting the motion vector.
  • the correction target pixel selection unit 46 extracts pixels whose maximum depth difference absolute value calculated by the depth information difference value detection unit 41 is equal to or greater than the first threshold Ta.
  • the first threshold value Ta is 32, for example, and the correction target pixel selection unit 46 extracts pixels whose maximum absolute value of the difference in depth information is 32 or more.
  • the pixels indicated by hatching in FIG. 9 are pixels having a maximum difference absolute value of depth information of 32 or more.
  • FIG. 10 is a diagram schematically illustrating an area including pixels having a maximum difference absolute value of depth information equal to or greater than the first threshold Ta.
  • the hatched area in FIG. 10 indicates an area in which pixels within a range of 7 ⁇ 7 pixels centering on a pixel having a maximum difference absolute value of depth information of 32 or more shown in FIG. 9 are extracted. .
  • pixels within a range of 7 ⁇ 7 pixels centering on a pixel having a maximum difference absolute value of depth information equal to or greater than the first threshold Ta are extracted.
  • the pixel is not particularly limited to this, and pixels within a range smaller than 7 ⁇ 7 pixels such as 3 ⁇ 3 pixels or 5 ⁇ 5 pixels may be extracted, and 7 ⁇ 7 pixels such as 9 ⁇ 9 pixels may be extracted. Pixels within a range larger than the pixels may be extracted.
  • the correction target pixel selection unit 46 includes an area including a pixel having the maximum absolute value of the difference in depth information equal to or greater than the first threshold Ta and a divided block having a block matching difference value equal to or greater than the second threshold Tc. Pixels in the overlapping region are selected as correction target pixels.
  • FIG. 11 is a diagram schematically illustrating a correction target pixel according to the first embodiment of the present invention. The areas indicated by hatching in FIG. 11 are areas including pixels having the maximum absolute value of the difference in depth information equal to or greater than the first threshold Ta shown in FIG. 10 and blocks equal to or greater than the second threshold Tc shown in FIG. A region overlapping with a divided block having a matching difference value is shown. In this way, the correction target pixel is selected.
  • step S14 the search block specifying unit 471 specifies a target correction target pixel from the correction target pixels selected by the correction target pixel selection unit 46.
  • step S15 the search block specifying unit 471 specifies search divided blocks that exist in the direction in which the maximum value of the difference absolute value of the depth information decreases from the specified target correction target pixel.
  • FIG. 12 is a diagram for describing a method for specifying a search division block according to the first embodiment of this invention.
  • the search block specifying unit 471 selects the first of the eight adjacent divided blocks (upper, lower, left, right, upper left, upper right, lower left, and lower right divided blocks) adjacent to the divided block including the specified target correction target pixel.
  • An adjacent divided block including a pixel having a maximum absolute value of difference in depth information equal to or greater than a threshold Ta and a pixel having a maximum absolute value of difference in depth information equal to or greater than a first threshold Ta are pixels to be corrected. Are excluded from the search divided blocks.
  • the divided block H shown in FIG. 12 eight pixels having the maximum absolute value of the difference in depth information equal to or greater than the first threshold Ta are arranged on the upper side in the divided block H.
  • the divided block G, the divided block I, the divided block L, and the divided block N adjacent to the divided block H include pixels having the maximum difference absolute value of depth information equal to or greater than the first threshold Ta.
  • the lower eight correction target pixels in the divided block H are respectively specified as the target correction target pixels
  • the search block specifying unit 471 specifies one of the lower eight correction target pixels in the divided block H as the target correction target pixel, the adjacent divided block B, divided block C, and divided block D are identified.
  • the divided block G, the divided block I, the divided block L, and the divided block N are excluded from the search divided blocks, and the adjacent divided block M is specified as the search divided block.
  • the divided block L shown in FIG. 12 eight pixels having the maximum difference absolute value of the depth information equal to or greater than the first threshold Ta are arranged on the left side in the divided block L.
  • the divided block G, the divided block H, the divided block Q, and the divided block R adjacent to the divided block L include pixels having the maximum absolute value of the difference in depth information equal to or greater than the first threshold Ta.
  • the search block specifying unit 471 specifies one of the right eight correction target pixels in the divided block L as the target correction target pixel, the adjacent divided block F, the divided block G, the divided block H, The divided block K, the divided block P, the divided block Q, and the divided block R are excluded from the search divided blocks, and the adjacent divided block M is specified as the search divided block.
  • the pixels having the maximum absolute value of the depth information difference equal to or greater than the first threshold Ta are arranged so as to divide the divided block S into the upper left side and the lower right side. Yes.
  • the divided block N and the divided block R adjacent to the divided block S include pixels having the maximum difference absolute value of depth information equal to or greater than the first threshold Ta.
  • the divided block O, the divided block T, the divided block W, the divided block X, and the divided block adjacent to the divided block S Between Y and the target correction target pixel, there is a pixel having a maximum difference absolute value of depth information equal to or greater than the first threshold Ta.
  • the divided blocks M adjacent to the divided block S and each target correction target pixel are There is a pixel having a maximum absolute value of difference in depth information equal to or greater than one threshold Ta.
  • the search block specifying unit 471 specifies one upper left correction target pixel in the divided block S as the target correction target pixel, the adjacent divided block N, divided block O, divided block R, divided block T, The divided block W, the divided block X, and the divided block Y are excluded from the search divided blocks, and the adjacent divided block M is specified as the search divided block.
  • the search block specifying unit 471 specifies one of the ten correction target pixels on the lower right side in the divided block S as the target correction target pixel, the adjacent divided block M, the divided block N, and the divided block Block R is excluded from the search divided blocks, and adjacent divided block O, divided block T, divided block W, divided block X, and divided block Y are specified as search divided blocks.
  • the search block specifying unit 471 specifies search divided blocks for correction target pixels in the other divided blocks G, I, M, N, Q, and R, respectively.
  • the search block specifying unit 471 searches and divides a divided block closest to the target correction target pixel among the plurality of candidate divided blocks. Identify as a block.
  • the search block specifying unit 471 may specify a divided block with the highest motion vector accuracy among the plurality of candidate divided blocks as a search divided block. More preferably, the search block specifying unit 471 specifies a search divided block in consideration of both the positional relationship between the target correction target pixel and the candidate divided block and the accuracy of the motion vector among the plurality of candidate divided blocks. May be.
  • the search block specifying unit 471 extracts a candidate divided block closest to the target correction target pixel from among a plurality of candidate divided blocks, and when a plurality of candidate divided blocks closest to the target correction target pixel are further extracted, Of the plurality of candidate divided blocks closest to the correction target pixel, the candidate divided block having the highest motion vector accuracy may be specified as the search divided block.
  • step S16 the second comparison unit 474 determines whether or not the block matching difference value of the search divided block is equal to or less than the third threshold value Ts.
  • the selection determination unit 484 causes the current divided block in which the target correction target pixel exists to exist. Are selected as the motion vector corresponding to the target correction target pixel. Specifically, the selection determination unit 484 selects the motion vector of the current divided block in which the target correction target pixel selected by the first motion vector selection unit 481 exists and the second motion vector selection unit 482. The switch 483 is switched so as to select the motion vector of the current divided block among the motion vectors of the search divided blocks.
  • the selection determining unit 484 is specified by the search block specifying unit 471 in step S18.
  • the motion vector of the search divided block is selected as the motion vector corresponding to the target correction target pixel.
  • the selection determination unit 484 selects the motion vector of the current divided block in which the target correction target pixel selected by the first motion vector selection unit 481 exists and the second motion vector selection unit 482.
  • the switch 483 is switched so that the motion vector of the search division block is selected from the motion vectors of the search division block. In this way, the selection determining unit 484 changes the motion vector of the target correction target pixel to the motion vector of the search division block.
  • step S19 the selection determining unit 484 determines whether or not the motion vector selection has been completed for all the correction target pixels.
  • the process returns to the process of step S14, and the search block specifying unit 471 causes the correction target pixel selection unit 46 to return.
  • the target correction target pixel for which the selection of the motion vector has not been completed is identified from the correction target pixels selected by the above. Of the pixels constituting the frame image, for the pixels other than the correction target pixel, the motion vector of the divided block to which each pixel belongs is selected.
  • step S19 if it is determined that the motion vector selection has been completed for all the correction target pixels (YES in step S19), the motion vector correction process is terminated, and the process proceeds to step S5 in FIG.
  • FIG. 13 is a diagram schematically showing a corrected motion vector according to the first embodiment of the present invention.
  • the motion vectors of the pixels in the divided blocks are changed.
  • the motion vector of the pixel having the maximum difference absolute value of the depth information equal to or greater than the first threshold Ta indicated by the hatching of each divided block G, I, Q, S is the same motion vector ( ⁇ 40, 32), and the motion vectors of the other pixels of the divided blocks G, I, Q, and S are changed to motion vectors (0, 0) different from those before the correction.
  • the corrected motion vector distribution shown in FIG. 13 approximates the shapes of the foreground image SG and the background image BG.
  • the foreground image SG and the background image BG are clearly distinguished, and the motion vector of the pixel at the boundary between the foreground image SG and the background image BG is corrected, so that the motion vector can be detected with high accuracy.
  • image quality can be prevented from deteriorating.
  • step S5 the video signal processing unit 5 performs predetermined image processing on the video signal using the motion vector corrected by the motion vector correction unit 4.
  • step S6 the display unit 6 displays the video signal that has been subjected to the predetermined image processing by the video signal processing unit 5.
  • the block matching threshold value setting unit 44 sets the second threshold value so that the value decreases as the maximum absolute value of the depth information difference of each pixel calculated by the depth information difference value detection unit 41 increases. Tc may be set.
  • the block matching threshold value setting unit 44 sets the maximum difference absolute value of the depth information of each pixel calculated by the depth information difference value detection unit 41 to 64 (for example, the first value). It is determined whether or not it is equal to or greater than a fourth threshold value greater than a threshold value Ta of 1.
  • the block matching threshold value setting unit 44 includes the target divided block and a plurality of divided blocks around the target divided block. The coefficient k multiplied by the average value of the block matching difference values in the region to be linearly changed so as to decrease as the maximum absolute value of the depth information difference increases. For example, the block matching threshold value setting unit 44 linearly changes the value of the coefficient k from 1.2 to 1.1.
  • the maximum difference absolute value of the depth information is large, the number of pixels selected as correction targets can be increased, and the motion vector can be corrected more reliably.
  • the block search unit 47 determines whether or not the maximum value of the absolute difference value of the depth information of each pixel calculated by the depth information difference value detection unit 41 is equal to or greater than a predetermined fourth threshold value. If it is determined that the maximum absolute value of the difference in depth information of each pixel is equal to or greater than a predetermined fourth threshold, the search range, that is, the number of divided blocks to be searched or the area of the search range is increased. May be.
  • the block search unit 47 sets the maximum difference absolute value of the depth information of each pixel calculated by the depth information difference value detection unit 41 to 64 (first value, for example). It is determined whether or not it is equal to or greater than a fourth threshold value that is greater than the threshold value Ta.
  • the block search unit 47 uses eight divided blocks (attention correction) adjacent to the divided block including the identified target correction target pixel. In addition to the 3 ⁇ 3 divided blocks excluding the divided block including the target pixel), a search divided block is selected from the 16 divided blocks further adjacent to the eight divided blocks adjacent to the divided block including the target correction target pixel. Identify.
  • the block search unit 47 includes 24 divided blocks (attention correction) around the divided block including the identified target correction target pixel.
  • the search divided block is specified from among 5 ⁇ 5 divided blocks excluding the divided block including the target pixel.
  • the block search unit 47 determines the 48 divided blocks around the divided block including the identified target correction target pixel (target correction).
  • a search divided block may be specified from among (7 ⁇ 7 divided blocks excluding divided blocks including the target pixel).
  • the block search unit 47 may increase the search range as the maximum difference absolute value of the depth information increases when it is determined that the maximum difference absolute value of the depth information is, for example, 64 or more. Good.
  • the motion vector of the target correction target pixel is corrected with higher accuracy by increasing the number of candidate divided blocks when specifying the search divided block. be able to.
  • the search block threshold value setting unit 473 determines whether or not the maximum value of the absolute difference value of the depth information of each pixel calculated by the depth information difference value detection unit 41 is greater than or equal to a predetermined fourth threshold value. If it is determined that the maximum difference absolute value of the depth information of each pixel is greater than or equal to a predetermined fourth threshold, the third threshold Ts may be increased.
  • the search block threshold value setting unit 473 has a maximum difference absolute value of the depth information of each pixel calculated by the depth information difference value detection unit 41, for example, 64 (first It is determined whether or not it is equal to or greater than a fourth threshold value greater than a threshold value Ta of 1.
  • the search block threshold value setting unit 473 is smaller than the second threshold value Tc and the difference absolute value per pixel is, for example, The third threshold value Ts is set to be 4. At this time, the third threshold value Ts is 64 when one divided block includes 16 pixels.
  • the motion vector of the target correction target pixel is changed by increasing the third threshold value Ts for comparison with the block matching difference value of the search divided block. Can be relaxed.
  • the block matching threshold value setting unit 44 may set the second threshold value Tc so that the value decreases as the frequency of each pixel detected by the frequency detection unit 42 increases.
  • the frequency detector 42 includes a high-pass filter having any of the characteristics shown in FIGS. 14 (A) to 14 (C).
  • FIG. 14A is a diagram illustrating an example of a high-pass filter having a characteristic (weighting coefficient) in the horizontal direction
  • FIG. 14B illustrates an example of a high-pass filter having a characteristic (weighting coefficient) in the vertical direction
  • FIG. 14C is a diagram illustrating an example of a high-pass filter having characteristics (weighting coefficients) in the horizontal and vertical directions.
  • the block matching threshold value setting unit 44 determines whether or not the absolute value of the output of the high-pass filter is, for example, 16 (fifth threshold value) or more. Judging.
  • the block matching threshold setting unit 44 is an area configured by the target divided block and a plurality of divided blocks around the target divided block.
  • the coefficient k multiplied by the average value of the block matching difference values is linearly changed so as to decrease as the absolute value of the output of the high-pass filter increases.
  • the block matching threshold value setting unit 44 linearly changes the value of the coefficient k from 1.2 to 1.1.
  • the absolute value of the output of the high-pass filter is large, that is, when the frequency of the pixel is large, the number of divided blocks selected as a correction target can be increased, and the motion vector can be more reliably corrected.
  • the block search unit 47 determines whether or not the frequency of each pixel detected by the frequency detection unit 42 is equal to or higher than a predetermined fifth threshold, and the frequency of each pixel is set to a predetermined first frequency.
  • the search range that is, the number of divided blocks to be searched or the area of the search range may be increased.
  • the block search unit 47 determines whether or not the absolute value of the output of the high-pass filter is, for example, 16 (fifth threshold value) or more. To do.
  • the block search unit 47 uses eight divided blocks (attention correction target pixels) adjacent to the divided block including the specified attention correction target pixel.
  • the search divided block is identified from the 16 divided blocks further adjacent to the eight divided blocks adjacent to the divided block including the target correction target pixel. .
  • the block search unit 47 includes 24 divided blocks (attention correction target pixels) around the divided block including the specified target correction target pixel.
  • the search divided block is identified from among the 5 ⁇ 5 divided blocks excluding the divided blocks including.
  • the block search unit 47 determines the 48 divided blocks (attention correction target pixels) around the divided block including the specified target correction target pixel.
  • the search divided block may be specified from among (7 ⁇ 7 divided blocks excluding divided blocks including). Further, the block search unit 47 may increase the search range as the absolute value of the output of the high-pass filter increases when it is determined that the absolute value of the output of the high-pass filter is, for example, 16 or more.
  • the target correction target pixel can be corrected with higher accuracy by increasing the number of candidate divided blocks when specifying the search divided block. Can be corrected.
  • the search block threshold value setting unit 473 determines whether the frequency of each pixel detected by the frequency detection unit 42 is equal to or higher than a predetermined fifth threshold value, and the frequency of each pixel is predetermined.
  • the third threshold value Ts may be increased when it is determined that the value is equal to or greater than the fifth threshold value.
  • the search block threshold value setting unit 473 determines whether or not the absolute value of the output of the high-pass filter is, for example, 16 (fifth threshold value) or more. Judging.
  • the search block threshold value setting unit 473 is smaller than the second threshold value Tc and the difference absolute value per pixel is, for example, 4.
  • the third threshold value Ts is set so that At this time, the third threshold value Ts is 64 when one divided block includes 16 pixels.
  • the target threshold value to be corrected is increased by increasing the third threshold value Ts for comparison with the block matching difference value of the search divided block.
  • the condition for changing the pixel motion vector can be relaxed.
  • the correction target pixel selection unit 46 determines whether or not the frequency of each pixel detected by the frequency detection unit 42 is equal to or higher than a predetermined fifth threshold, and the frequency of each pixel is predetermined. If it is determined that the value is equal to or greater than the fifth threshold value, the first threshold value Ta may be decreased.
  • the correction target pixel selection unit 46 determines whether or not the absolute value of the output of the high-pass filter is, for example, 16 (fifth threshold value) or more. Judging.
  • the correction target pixel selection unit 46 changes the first threshold Ta from 32 to 8. That is, the correction target pixel selection unit 46 determines whether or not the calculated maximum absolute value of the difference in the depth information of each pixel is 8 or more.
  • the absolute value of the output of the high-pass filter is large, that is, when the frequency of the pixel is large, the number of pixels selected as a correction target can be increased, and the motion vector can be corrected more reliably.
  • FIG. 15 is a diagram for explaining an additional process for setting the first threshold Ta based on the frequency of each pixel.
  • a frame image FG2 shown in FIG. 15 includes a first background image BG1 representing the sky, a second background image BG2 representing a tree existing in front of the first background image BG1, and a first background representing a bus. It includes a foreground image B1 and a second foreground image B2 representing a bus existing on the near side of the first foreground image B1.
  • the deterioration of the image quality is conspicuous because it is not a monotonous pattern such as the sky, that is, an image having a low frequency, but a complex such as a tree or a bus It is the one with a higher pattern, that is, a higher frequency image.
  • the frequency of each pixel is large, by reducing the first threshold Ta, the number of pixels selected as correction targets can be increased, and the motion vector can be corrected more reliably.
  • a liquid crystal display device is described as an example of a video display device.
  • the present invention is not particularly limited to this, and one field or one frame is divided into a plurality of subfields.
  • the present invention can be similarly applied to a plasma display device that performs gradation display.
  • the video signal processing unit 5 and the display unit 6 have the following functions.
  • the video signal processing unit 5 divides one field or one frame into a plurality of subfields, and combines the light-emitting subfields that emit light and the non-light-emitting subfields that do not emit light to perform gradation display. Convert to flash data.
  • the video signal processing unit 5 sequentially converts one frame image data, that is, one field of image data into light emission data of each subfield.
  • One field is composed of K subfields (where K is an integer equal to or greater than 2), and each subfield is given a predetermined weight corresponding to the luminance, and the luminance of each subfield changes according to this weighting.
  • the light emission period is set to For example, when 7 subfields are used and weighting of 2 7 is performed, the weights of the first to seventh subfields are 1, 2, 4, 8, 16, 32, 64, respectively.
  • the video signal processing unit 5 spatially rearranges the converted emission data of each subfield for each pixel of the frame N according to the motion vector corrected by the motion vector correction unit 4, thereby Rearranged light emission data of each subfield is generated for each pixel and output to the display unit 6.
  • the video signal processing unit 5 identifies the subfield that emits light among the subfields of each pixel of the frame N, and moves the motion vector so that the temporally preceding subfield moves greatly according to the subfield arrangement order.
  • the light emission data of the subfield corresponding to the pixel at the position spatially moved backward by the number of pixels corresponding to is changed to the light emission data of the subfield of the pixel before the movement.
  • the subfield rearrangement method is not particularly limited to this example.
  • the subfield rearrangement method is spatially equivalent to the pixel corresponding to the motion vector so that the temporally preceding subfield moves greatly according to the subfield arrangement order.
  • the display unit 6 includes a plasma display panel, a panel drive circuit, and the like, and controls on / off of each subfield of each pixel of the plasma display panel based on the light emission data rearranged by the video signal processing unit 5. Display moving images.
  • FIG. 16 is a diagram showing a configuration of a video display device according to the second embodiment of the present invention.
  • 16 includes a video signal input unit 1, a motion vector detection unit 3, a motion vector correction unit 4, a video signal processing unit 5, a display unit 6, and a depth information acquisition unit 7.
  • the video processing apparatus includes a video signal input unit 1, a motion vector detection unit 3, a motion vector correction unit 4, a video signal processing unit 5, and a depth information acquisition unit 7.
  • the 3D video signal is input to the video signal input unit 1 and the depth information acquisition unit 7.
  • the three-dimensional video signal includes a left-eye video signal created to be viewed with the left eye and a right-eye video signal created to be viewed with the right eye.
  • the right-eye video signal and the left-eye video signal have parallax for displaying the video in a three-dimensional manner.
  • the depth information acquisition unit 7 detects depth information based on the right-eye video signal and the left-eye video signal.
  • the depth information acquisition unit 7 detects the depth information of each pixel based on the shift amount between the pixels constituting the right-eye video signal and the pixels constituting the left-eye video signal. Note that the depth information detection method using the right-eye video signal and the left-eye video signal is a known technique, and thus detailed description thereof is omitted.
  • the display unit 6 displays the right-eye video signal and the left-eye video signal by alternately switching in time.
  • the right-eye video signal and the left-eye video signal are alternately switched in time and displayed at the same time.
  • the left and right fields of view are temporally separated using glasses that close the field of view, and a barrier or lens that limits the display angle of the image displayed on the display in front of the display unit 6 is used to separate the left and right eyes.
  • Various methods such as a method of displaying a right-eye video signal and a left-eye video signal can be applied.
  • the operation of the video display device 11 in the second embodiment is the same as that of the video display device 10 in the first embodiment shown in FIG. 3, and the motion vector correction unit in the second embodiment.
  • the operation of 4 is the same as the operation of the motion vector correction unit 4 in the first embodiment shown in FIG.
  • the depth information is not input and the 3D video signal including the right-eye video signal and the left-eye video signal having parallax is input, It is possible to detect a motion vector with high accuracy and to prevent deterioration in image quality.
  • FIG. 17 is a diagram showing a configuration of a video display device according to the third embodiment of the present invention.
  • the video display device 12 illustrated in FIG. 17 includes a video signal input unit 1, a motion vector detection unit 3, a motion vector correction unit 4, a video signal processing unit 5, a display unit 6, and a depth information acquisition unit 8. Note that the same reference numerals are given to the same components as those of the video display device in the first embodiment, and description thereof is omitted.
  • the video processing apparatus includes a video signal input unit 1, a motion vector detection unit 3, a motion vector correction unit 4, a video signal processing unit 5, and a depth information acquisition unit 8.
  • the two-dimensional video signal is input to the video signal input unit 1 and the depth information acquisition unit 8.
  • the two-dimensional video signal is a video signal composed of a plurality of temporally continuous frame images.
  • the depth information acquisition unit 8 detects depth information based on the saturation of each pixel constituting the current frame image. Generally, the scenery has a tendency that the saturation decreases as the distance increases, and the saturation increases as it approaches. Therefore, the depth information acquisition unit 8 detects the saturation of each pixel constituting the current frame image, and uses the detected saturation as the depth information. Since a method for detecting depth information from a two-dimensional video signal is a known technique, detailed description thereof is omitted.
  • the display unit 6 may display the two-dimensional video signal as it is, but generates a left-eye video signal and a right-eye video signal from the two-dimensional video signal based on the depth information detected by the depth information acquisition unit 8,
  • the right-eye video signal and the left-eye video signal may be switched alternately in time and displayed. In this case, an image can be displayed in three dimensions.
  • the operation of the video display device 12 in the third embodiment is the same as the operation of the video display device 10 in the first embodiment shown in FIG. 3, and the motion vector correction unit in the third embodiment.
  • the operation of 4 is the same as the operation of the motion vector correction unit 4 in the first embodiment shown in FIG.
  • the motion vector can be detected with high accuracy as in the first embodiment, and the image quality can be improved. Deterioration can be prevented.
  • FIG. 18 is a diagram showing a configuration of a motion vector correction unit of the video display device according to the fourth embodiment of the present invention.
  • the configuration other than the motion vector correction unit of the video display device according to the fourth embodiment is the same as the configuration of the video display device 10 according to the first embodiment, and thus description thereof is omitted.
  • the motion vector correction unit 4 illustrated in FIG. 18 includes a depth information difference value detection unit 41, a frequency detection unit 42, a first block matching difference value selection unit 43, a block matching threshold value setting unit 44, a first comparison unit 45, and a correction.
  • a target block selection unit 49, a block search unit 50, and a motion vector change unit 51 are provided.
  • symbol is attached
  • the correction target block selection unit 49 has a maximum absolute value of the difference in depth information of each pixel calculated by the depth information difference value detection unit 41 equal to or greater than a predetermined first threshold Ta, and the motion vector detection unit 3 A small block including a pixel whose calculated block matching difference value is equal to or greater than a predetermined second threshold Tc is selected as a correction target block for correcting a motion vector.
  • the divided block includes a plurality of small blocks including a plurality of pixels.
  • the number of pixels constituting the small block is smaller than the number of pixels constituting the divided block.
  • a small block includes 2 ⁇ 2 pixels.
  • the block search unit 50 calculates the block matching difference value from the target correction target block among the correction target blocks selected by the correction target block selection unit 49 in a direction in which the maximum absolute value of the difference in depth information decreases.
  • a divided block that is equal to or less than a predetermined third threshold value Ts is searched for within a predetermined search range for each correction target block.
  • the block search unit 50 includes a search block identification unit 501, a second block matching difference value selection unit 472, a search block threshold value setting unit 502, and a second comparison unit 474.
  • the search block specifying unit 501 specifies a search divided block that exists in a direction in which the maximum value of the difference absolute value of the depth information decreases from the target correction target block among the correction target blocks selected by the correction target block selecting unit 49. To do.
  • the search block threshold value setting unit 502 sets a third threshold value Ts for comparison with the block matching difference value of the search divided block.
  • the motion vector change unit 51 finds the motion vector of each pixel included in the target correction target block by the block search unit 50. Change to block motion vector.
  • the motion vector changing unit 51 does not change the motion vector of the target correction target pixel when the block search unit 50 does not find a divided block equal to or smaller than the third threshold Ts.
  • the motion vector change unit 51 includes a first motion vector selection unit 481, a second motion vector selection unit 511, a switch 512, and a selection determination unit 513.
  • the second motion vector selection unit 511 selects the motion vector of the search divided block specified by the search block specification unit 501 and outputs it to the switch 483.
  • the switch 512 switches between the motion vector of the current divided block selected by the first motion vector selection unit 481 and the motion vector of the search divided block selected by the second motion vector selection unit 511.
  • the selection determination unit 513 causes the switch 512 to output the motion vector of the search division block.
  • the motion vector of the switching divided search block is output as the motion vector of the target correction target pixel.
  • the selection determination unit 513 determines the search divided block when the second comparison unit 474 determines that the block matching difference value of the search divided block is larger than the third threshold Ts, or the search block specifying unit 501 specifies the search divided block. If not, the switch 512 is switched to output the motion vector of the current divided block, and the motion vector of the current divided block is output as the motion vector of the target correction target pixel.
  • the selection determination unit 513 selects either the motion vector of the current divided block or the motion vector of the search divided block for the correction target pixel among the pixels constituting the current frame image, and selects other than the correction target pixel. For the pixel, the motion vector of the current divided block is selected.
  • the motion vector detection unit 3 corresponds to an example of a motion vector detection unit
  • the motion vector correction unit 4 corresponds to an example of a motion vector correction unit
  • the depth information difference value detection unit 41 includes
  • the correction target block selection unit 49 corresponds to an example of a correction target block selection unit
  • the block search unit 50 corresponds to an example of a block search unit
  • the motion vector change unit 51 corresponds to an example of a depth information difference value calculation unit.
  • the search block threshold value setting unit 502 corresponds to an example of a motion vector change unit
  • the search threshold value setting unit 502 corresponds to an example of a first threshold value setting unit and a third threshold value setting unit
  • the block matching threshold value setting unit 44 is set to a second threshold value setting unit and This corresponds to an example of a fourth threshold setting unit.
  • FIG. 19 is a flowchart for explaining a motion vector correction process according to the fourth embodiment of the present invention.
  • steps S21 and S22 is the same as the processing in steps S11 and S12 shown in FIG.
  • the correction target block selection unit 49 has the maximum absolute value of the difference in depth information of each pixel calculated by the depth information difference value detection unit 41 equal to or greater than a predetermined first threshold value Ta, and a motion vector.
  • a small block including a pixel whose block matching difference value calculated by the detection unit 3 is equal to or larger than a predetermined second threshold Tc is selected as a correction target block for correcting a motion vector.
  • the correction target block selection unit 49 extracts a small block including x% or more pixels whose maximum absolute value of the difference in depth information calculated by the depth information difference value detection unit 41 is equal to or more than the first threshold Ta.
  • the first threshold Ta is, for example, 32
  • x% is, for example, 75%.
  • the correction target block selection unit 49 extracts a small block including 75% or more of pixels having a maximum difference absolute value of depth information of 32 or more.
  • FIG. 20 is a diagram schematically showing the maximum value of the absolute difference value of the depth information in the fourth embodiment of the present invention.
  • the maximum difference absolute value of depth information shown in FIG. 20 is calculated based on the depth information shown in FIG. 5, and is the same as the maximum difference absolute value of depth information shown in FIG.
  • a small block indicated by hatching in FIG. 20 is a small block including 75% or more of pixels having a maximum difference absolute value of depth information of 32 or more.
  • one divided block is composed of, for example, 16 (4 ⁇ 4) pixels
  • one small block is composed of, for example, 4 (2 ⁇ 2) pixels.
  • FIG. 20 is a diagram schematically illustrating an area including a small block including 75% or more of pixels having a maximum difference absolute value of depth information equal to or greater than the first threshold Ta.
  • the hatched area in FIG. 21 is a small area within the range of a 3 ⁇ 3 small block centered on a small block including 75% or more of pixels whose maximum absolute value of the difference in depth information shown in FIG. 20 is 32 or more. The area from which the block was extracted is shown.
  • the motion vector should be corrected by extracting a small block in a wider range than the small block including 75% or more of pixels having the maximum difference absolute value of depth information equal to or greater than the first threshold Ta. It is possible to prevent omission of detection of small blocks.
  • the maximum value of the absolute value of the difference in depth information is within the range of a 3 ⁇ 3 small block centered on a small block that includes 75% or more of pixels that are equal to or greater than the first threshold Ta.
  • the small block is extracted, the present invention is not particularly limited to this, and a small block within a range larger than a 3 ⁇ 3 small block such as a 5 ⁇ 5 small block may be extracted.
  • the correction target block selection unit 49 includes a region including a small block including x% or more of pixels having a maximum absolute value of difference in depth information equal to or greater than the first threshold Ta, and a block matching difference equal to or greater than the second threshold Tc. A pixel in a region overlapping with a divided block having a value is selected as a correction target block.
  • FIG. 22 is a diagram schematically illustrating a correction target block according to the fourth embodiment of the present invention. The areas indicated by hatching in FIG. 22 include an area including a small block including x% or more of pixels having the maximum absolute value of the difference in depth information equal to or greater than the first threshold Ta illustrated in FIG. 21 and the second illustrated in FIG.
  • An overlapping area is shown with a divided block having a block matching difference value equal to or greater than the threshold value Tc. In this way, the correction target block is selected.
  • the area surrounded by the thick line in FIG. 22 represents an area including a small block including x% or more of pixels having the maximum difference absolute value of depth information equal to or greater than the first threshold Ta shown in FIG. .
  • step S24 the search block specifying unit 501 specifies a target correction target block from the correction target blocks selected by the correction target block selecting unit 49.
  • step S25 the search block specifying unit 501 specifies a search divided block that exists in a direction in which the maximum value of the difference absolute value of the depth information decreases from the specified target correction target block.
  • FIG. 23 is a diagram for explaining a method for specifying a search division block according to the fourth embodiment of the present invention.
  • the search block specifying unit 501 is the first of eight adjacent divided blocks (upper, lower, left, right, upper left, upper right, lower left, and lower right divided blocks) adjacent to the divided block including the identified target correction target block.
  • An adjacent divided block in which a small block including x% or more pixels exists between the target correction target block is excluded from the search divided blocks.
  • two small blocks including x% or more of pixels having the maximum absolute value of the difference in depth information equal to or greater than the first threshold Ta are arranged on the upper side in the divided block H.
  • the divided block G, the divided block I, the divided block L, and the divided block N adjacent to the divided block H there are small blocks including x% or more of pixels having the maximum absolute value of the difference in depth information that is equal to or greater than the first threshold Ta. included.
  • the divided block B, the divided block C, the divided block D adjacent to the divided block H, and each target correction target block Are small blocks including x% or more of pixels having the maximum absolute value of the difference in depth information equal to or greater than the first threshold Ta.
  • the search block specifying unit 501 adjoins the divided block B, the divided block C, and the divided block D.
  • the divided block G, the divided block I, the divided block L, and the divided block N are excluded from the search divided blocks, and the adjacent divided block M is specified as the search divided block.
  • two small blocks including x% or more of pixels having the maximum absolute value of the difference in depth information equal to or greater than the first threshold value Ta are arranged on the left side in the divided block L.
  • the divided block G, the divided block H, the divided block Q, and the divided block R that are adjacent to the divided block L include small blocks that include x% or more pixels having the maximum difference absolute value of depth information that is equal to or greater than the first threshold Ta. included.
  • the search block specifying unit 501 specifies one of the right two correction target blocks in the divided block L as the target correction target block, the adjacent divided block F, the divided block G, the divided block H, The divided block K, the divided block P, the divided block Q, and the divided block R are excluded from the search divided blocks, and the adjacent divided block M is specified as the search divided block.
  • the divided block S shown in FIG. 23 there is one small block on the upper left side in the divided block S that includes x% or more of pixels having the maximum absolute value of the difference in depth information equal to or greater than the first threshold Ta.
  • Each of the divided block N and the divided block R adjacent to the divided block S includes a small block including x% or more of pixels having the maximum difference absolute value of the depth information equal to or greater than the first threshold Ta.
  • the divided block M adjacent to the divided block S and each target correction target block are In between, there is a small block including x% or more of pixels having the maximum absolute value of the difference in depth information equal to or greater than the first threshold Ta.
  • the search block specifying unit 501 specifies one of the three correction target blocks other than the one upper correction target block in the divided block S as the target correction target block, the adjacent divided block M, The divided block N and the divided block R are excluded from the search divided blocks, and the adjacent divided block O, divided block T, divided block W, divided block X, and divided block Y are specified as the search divided blocks.
  • the search block specifying unit 501 specifies search divided blocks for correction target blocks in other divided blocks G, I, M, N, Q, and R, respectively.
  • the search block specifying unit 501 searches and divides the divided block closest to the target correction target block among the plurality of candidate divided blocks. Identifies as a block.
  • the search block specifying unit 501 may specify a divided block having the highest motion vector accuracy among the plurality of candidate divided blocks as a search divided block. More preferably, the search block specifying unit 501 specifies a search divided block in consideration of both the positional relationship between the target correction target block and the candidate divided block and the accuracy of the motion vector among the plurality of candidate divided blocks. May be.
  • the search block specifying unit 501 extracts a candidate divided block closest to the target correction target block from among a plurality of candidate divided blocks, and when a plurality of candidate divided blocks closest to the target correction target block are further extracted, Of the plurality of candidate divided blocks closest to the correction target block, the candidate divided block with the highest motion vector accuracy may be specified as the search divided block.
  • step S26 is the same as the processing in step S16 shown in FIG.
  • step S27 the selection determination unit 513 causes the current divided block in which the target correction target block exists to exist. Is selected as a motion vector corresponding to the target correction target block. Specifically, the selection determination unit 513 selects the motion vector of the current divided block in which the target correction target block selected by the first motion vector selection unit 481 exists and the second motion vector selection unit 511. The switch 512 is switched so that the motion vector of the current divided block is selected from the motion vectors of the search divided blocks.
  • the selection determination unit 513 is specified by the search block specifying unit 501 in step S28.
  • the motion vector of the search divided block is selected as the motion vector corresponding to the target correction target block.
  • the selection determination unit 513 selects the motion vector of the current divided block in which the target correction target block selected by the first motion vector selection unit 481 exists and the second motion vector selection unit 511.
  • the switch 512 is switched so that the motion vector of the search divided block is selected from the motion vectors of the search divided blocks. In this way, the selection determining unit 513 changes the motion vector of the target correction target block to the motion vector of the search divided block.
  • step S29 the selection determining unit 513 determines whether or not the motion vector selection has been completed for all the correction target blocks.
  • the process returns to step S24, and the search block specifying unit 501 includes the correction target block selection unit 49.
  • the target correction target block for which the selection of the motion vector has not been completed is identified from the correction target blocks selected by the above.
  • the motion vector of the divided block to which each small block belongs is selected.
  • step S29 when it is determined that the motion vector selection has been completed for all the correction target blocks (YES in step S29), the motion vector correction process is terminated, and the process proceeds to step S5 in FIG.
  • FIG. 24 is a diagram schematically showing a corrected motion vector according to the fourth embodiment of the present invention.
  • the motion vector of the small block in the divided block is changed.
  • the motion vector of the small block including x% or more of pixels having the maximum absolute value of the difference in depth information equal to or greater than the first threshold Ta, indicated by hatching of each of the divided blocks G, I, Q, and S, is The same motion vector ( ⁇ 40, 32), and the motion vectors of the other small blocks of each divided block G, I, Q, S are changed to a motion vector (0, 0) different from that before the correction.
  • the corrected motion vector distribution shown in FIG. 24 approximates the shapes of the foreground image SG and the background image BG.
  • the foreground image SG and the background image BG are clearly distinguished, and the foreground image Since the motion vector of the pixel at the boundary portion between the SG and the background image BG is corrected, it is possible to detect the motion vector with high accuracy and to prevent deterioration of the image quality.
  • the second threshold value Tc may be set so that
  • the block matching threshold value setting unit 44 sets the maximum absolute value of the difference in depth information of 75% or more of the pixels constituting the small block, for example, It is determined whether or not it is equal to or greater than 64 (fourth threshold value greater than the first threshold value Ta).
  • 64 fourth threshold value greater than the first threshold value Ta
  • the block matching threshold setting unit 44 The maximum value of the difference absolute value of the depth information is increased by the coefficient k multiplied by the average value of the block matching difference values in the area composed of the divided block and the plurality of divided blocks around the target divided block. It is linearly changed so as to become smaller.
  • the block matching threshold value setting unit 44 linearly changes the value of the coefficient k from 1.2 to 1.1.
  • the small block includes a pixel having a large maximum difference absolute value of the depth information
  • the number of small blocks selected as the correction target can be increased, and the motion vector can be more reliably corrected. it can.
  • the search block specifying unit 501 When a value of 0 to 127 is input as the depth information, the search block specifying unit 501 has a maximum absolute value of difference in depth information of 75% or more of a plurality of pixels constituting the small block, for example, 64. It is determined whether or not (fourth threshold value larger than the first threshold value Ta) or more.
  • the search block specifying unit 501 specifies In addition to the eight divided blocks adjacent to the divided block including the target correction target block (3 ⁇ 3 divided blocks excluding the divided block including the target correction target block), 8 adjacent to the divided block including the target correction target block A search divided block is specified from 16 divided blocks further adjacent to one divided block.
  • the search block specifying unit 501 determines the specified attention when it is determined that the maximum difference absolute value of the depth information of 75% or more of the plurality of pixels constituting the small block is, for example, 64 or more.
  • a search divided block is specified from 24 divided blocks (5 ⁇ 5 divided blocks excluding the divided block including the target correction target block) around the divided block including the correction target block.
  • the search block specifying unit 501 determines the specified attention when it is determined that the maximum value of the absolute difference value of the depth information of 75% or more of the pixels constituting the small block is 64 or more, for example.
  • the search divided block may be specified from among the 48 divided blocks (7 ⁇ 7 divided blocks excluding the divided block including the target correction target block) around the divided block including the correction target block.
  • the search block specifying unit 501 determines that the maximum value of the absolute difference value of the depth information of 75% or more of the plurality of pixels constituting the small block is, for example, 64 or more. As the maximum difference absolute value increases, the search range may be increased.
  • the target correction target can be corrected more accurately by increasing the number of candidate divided blocks when specifying the search divided block.
  • the motion vector of the block can be corrected.
  • the search block threshold value setting unit 502 sets the maximum absolute value of the difference in depth information of 75% or more of a plurality of pixels constituting the small block as an example. It is determined whether or not it is equal to or greater than 64 (fourth threshold value greater than the first threshold value Ta).
  • 64 fourth threshold value greater than the first threshold value Ta
  • the search block threshold value setting unit 502 The third threshold value Ts is set so that the absolute value of difference per pixel is 4, for example, smaller than the threshold value Tc of 2. At this time, the third threshold value Ts is 64 when one divided block includes 16 pixels.
  • the attention correction is performed by increasing the third threshold value Ts for comparison with the block matching difference value of the search divided block.
  • the condition for changing the motion vector of the target block can be relaxed.
  • the block matching threshold value setting unit 44 determines that the target divided block is And the value of the coefficient k multiplied by the average value of the block matching difference values in the area composed of the divided block around the target divided block is 75% or more of the plurality of pixels constituting the small block.
  • the absolute value of the output of the high-pass filter of this pixel is linearly changed so as to decrease as the absolute value increases.
  • the block matching threshold value setting unit 44 linearly changes the value of the coefficient k from 1.2 to 1.1.
  • the small block includes pixels with a high absolute value of the output of the high-pass filter, that is, when the small block includes pixels with a high frequency
  • the number of divided blocks selected as correction targets can be increased.
  • the motion vector can be corrected more reliably.
  • a search range that is, a divided block to be searched Or the area of the search range may be increased.
  • the search block specifying unit 501 When a video signal having a luminance value of 0 to 255 for each pixel is input, the search block specifying unit 501 outputs the absolute value of the output of the high-pass filter for 75% or more of the plurality of pixels constituting the small block. It is determined whether or not the value is, for example, 16 (fifth threshold value) or more.
  • the search block specifying unit 501 specifies the specified attention correction.
  • the eight partitions adjacent to the divided block including the target correction target block A search divided block is specified from 16 divided blocks further adjacent to the block.
  • the search block specifying unit 501 specifies the specified target correction target A search divided block is specified from 24 divided blocks (5 ⁇ 5 divided blocks excluding the divided block including the target correction target block) around the divided block including the block.
  • the search block specifying unit 501 determines the specified attention correction target when the absolute value of the high-pass filter output of 75% or more of the plurality of pixels constituting the small block is, for example, 16 or more.
  • a search divided block may be specified from 48 divided blocks (7 ⁇ 7 divided blocks excluding the divided block including the target correction target block) around the divided blocks including the block.
  • the search block specifying unit 501 outputs the output of the high-pass filter. As the absolute value increases, the search range may be increased.
  • the small block includes pixels with a high absolute value of the output of the high-pass filter, that is, when the small block includes pixels with a high frequency
  • the number of candidate divided blocks when specifying the search divided block is increased. This makes it possible to correct the motion vector of the target correction target block with higher accuracy.
  • the search block threshold value setting unit 502 When a video signal having a luminance value of each pixel from 0 to 255 is input, the search block threshold value setting unit 502 outputs the output of the high-pass filter of 75% or more of the pixels constituting the small block. It is determined whether or not the absolute value is 16 (fifth threshold value) or more, for example.
  • the search block threshold value setting unit 502 The third threshold value Ts is set so that it is smaller than the threshold value Tc and the difference absolute value per pixel is 4, for example. At this time, the third threshold value Ts is 64 when one divided block includes 16 pixels.
  • the third block for comparison with the block matching difference value of the search divided block is used.
  • the correction target block selecting unit 49 When a video signal having a luminance value of 0 to 255 for each pixel is input, the correction target block selecting unit 49 outputs the output of the high-pass filter of 75% or more of the plurality of pixels constituting the small block. It is determined whether or not the absolute value is 16 (fifth threshold value) or more, for example.
  • the correction target block selection unit 49 selects the first block The threshold value Ta is changed from 32 to 8. That is, the correction target block selection unit 49 determines whether or not the maximum difference absolute value of the depth information of each pixel constituting the small block is 8 or more.
  • the small block includes a pixel having a large absolute value of the output of the high-pass filter, that is, when the small block includes a pixel having a large frequency
  • the number of small blocks selected as a correction target can be increased.
  • the motion vector can be corrected more reliably.
  • a liquid crystal display device is described as an example of an image display device.
  • the present invention is not particularly limited to this, and one field or one frame is divided into a plurality of subfields.
  • the present invention can be similarly applied to a plasma display device that performs gradation display.
  • the fourth embodiment even when the depth information is not input and a 3D video signal including a right-eye video signal and a left-eye video signal having parallax is input, Similar to the embodiment, it is possible to detect a motion vector with high accuracy and to prevent deterioration in image quality.
  • a motion vector can be detected with high accuracy as in the third embodiment.
  • image quality deterioration can be prevented.
  • a video display device detects a motion vector from a plurality of frame images in a video signal including a plurality of temporally continuous frame images, and uses the detected motion vectors to generate the video signal.
  • a video display device that performs predetermined image processing and displays, a motion vector detection unit that detects a motion vector of a current frame image using at least two or more frame images that are temporally changed, and the current
  • a depth information acquisition unit that acquires depth information representing a position in the depth direction of each pixel constituting the frame image, and the motion vector detection unit that is detected based on the depth information acquired by the depth information acquisition unit
  • a motion vector correction unit that corrects the motion vector.
  • the motion vector detection unit detects the motion vector of the current frame image using at least two or more frame images that are temporally changed.
  • the depth information acquisition unit acquires depth information indicating the position in the depth direction of each pixel constituting the current frame image. Then, the detected motion vector is corrected by the motion vector correction unit based on the acquired depth information.
  • the foreground image and the background image are clearly distinguished by the depth information representing the position in the depth direction, and the motion vector of the pixel at the boundary between the foreground image and the background image is corrected, so that the motion vector is detected with high accuracy. It is possible to prevent image quality deterioration.
  • the motion vector detection unit divides the frame image into a plurality of divided blocks including a plurality of pixels, detects the motion vector for each of the divided blocks, and detects the current frame.
  • the absolute value of the difference between the luminance value of each pixel in the target divided block of the image and the luminance value of each pixel of the reference divided block referred to when detecting the motion vector of the past frame image is calculated.
  • a total value of the pixel difference absolute value in the target divided block is calculated for each divided block as a block matching difference value of the target divided block, and the motion vector correction unit Depth for which the maximum value of the absolute difference between the depth information of the target pixel and the depth information of the adjacent pixel adjacent to the target pixel is calculated for each pixel
  • the maximum difference absolute value of the depth information of each pixel calculated by the report difference value calculation unit and the depth information difference value calculation unit is equal to or greater than a predetermined first threshold and is calculated by the motion vector detection unit.
  • a correction target pixel selection unit that selects a pixel whose block matching difference value is equal to or greater than a predetermined second threshold as a correction target pixel for correcting the motion vector, and the correction selected by the correction target pixel selection unit
  • a target block to be corrected is a divided block whose block matching difference value is equal to or smaller than a predetermined third threshold value in a direction in which the maximum value of the absolute difference value of the depth information decreases from the target correction target pixel among the target pixels.
  • a block search unit that searches within a predetermined search range every time, and a block that is equal to or less than the third threshold is found by the block search unit.
  • the motion vector detection unit divides the frame image into a plurality of divided blocks including a plurality of pixels, and a motion vector is detected for each of the divided blocks, and the current frame image in the target divided block is detected.
  • the absolute value of the difference between the luminance value of each pixel and the luminance value of each pixel of the reference divided block referenced when detecting the motion vector of the past frame image is calculated.
  • a total value in the target divided block is calculated for each divided block as a block matching difference value of the target divided block.
  • the depth information difference value calculation unit calculates, for each pixel, the maximum difference absolute value between the depth information of the target pixel of the pixels constituting the current frame image and the depth information of the adjacent pixel adjacent to the target pixel.
  • the maximum value of the difference absolute value of the depth information of each pixel calculated by the correction target pixel selection unit is greater than or equal to a predetermined first threshold, and the calculated block matching difference value is a predetermined second threshold.
  • the above pixel is selected as a correction target pixel for correcting the motion vector.
  • the block matching difference value is equal to or smaller than a predetermined third threshold value from the target correction target pixel among the correction target pixels selected by the block search unit in a direction in which the maximum value of the difference absolute value of the depth information decreases. A certain divided block is searched. When a divided block that is equal to or smaller than the third threshold is found, the motion vector changing unit changes the motion vector of the target correction target pixel to the motion vector of the found divided block.
  • a pixel that is included in a divided block with low motion vector reliability and has a large maximum absolute value of the difference in depth information of each pixel is selected as a correction target pixel. Therefore, a boundary portion between the foreground image and the background image Can be reliably selected as a correction target pixel.
  • the motion vector of the target correction target pixel is changed to the motion vector of the search divided block, a motion vector with low reliability can be corrected to a motion vector with high reliability.
  • the block search unit may determine whether the maximum absolute value of the difference in depth information of each pixel calculated by the depth information difference value calculation unit is equal to or greater than a predetermined fourth threshold value. It is preferable to increase the search range when it is determined that the maximum difference absolute value of the depth information of each pixel is equal to or greater than a predetermined fourth threshold value.
  • the block search unit determines whether or not the calculated maximum absolute value of the difference in the depth information of each pixel is equal to or greater than the predetermined fourth threshold, and the difference in the depth information of each pixel. When it is determined that the maximum absolute value is equal to or greater than the predetermined fourth threshold, the search range is increased.
  • the motion vector of the target correction target pixel can be corrected with higher accuracy by increasing the number of candidate divided blocks when specifying the search divided block. it can.
  • the motion vector correction unit may determine whether the maximum value of the absolute difference value of the depth information of each pixel calculated by the depth information difference value calculation unit is a predetermined fourth threshold value or more.
  • the first threshold setting unit determines whether or not the calculated maximum absolute value of the difference in the depth information of each pixel is greater than or equal to the predetermined fourth threshold, and the depth of each pixel. When it is determined that the maximum absolute value of the information difference is equal to or greater than a predetermined fourth threshold value, the third threshold value is increased.
  • the condition for changing the motion vector of the target correction target pixel is increased by increasing the third threshold value for comparison with the block matching difference value of the search divided block. Can be relaxed.
  • the motion vector correction unit may reduce the value as the maximum absolute value of the depth information difference of each pixel calculated by the depth information difference value calculation unit increases. It is preferable to further include a second threshold value setting unit that sets the second threshold value.
  • the second threshold value is set by the second threshold value setting unit so that the value becomes smaller as the maximum value of the absolute difference value of the calculated depth information of each pixel becomes larger.
  • the maximum difference absolute value of the depth information is large, the number of pixels selected as correction targets can be increased, and the motion vector can be corrected more reliably.
  • the video display device may further include a frequency detection unit that detects a frequency of each pixel of the current frame image, and the block search unit has a predetermined frequency of each pixel detected by the frequency detection unit. It is preferable to increase the search range when it is determined whether the frequency of each pixel is equal to or higher than a predetermined fifth threshold.
  • the frequency detection unit detects the frequency of each pixel of the current frame image.
  • the block search unit determines whether the frequency of each detected pixel is equal to or higher than a predetermined fifth threshold, and determines that the frequency of each pixel is higher than a predetermined fifth threshold The search range is increased.
  • the motion vector of the target correction target pixel can be corrected with higher accuracy by increasing the number of candidate divided blocks when specifying the search divided block.
  • the video display device may further include a frequency detection unit that detects a frequency of each pixel of the current frame image, and the motion vector correction unit has a predetermined frequency detected by the frequency detection unit.
  • a third threshold value setting unit for increasing the third threshold value when it is determined whether the frequency of each pixel is equal to or higher than a predetermined fifth threshold value; It is preferable to include.
  • the frequency detection unit detects the frequency of each pixel of the current frame image. Then, the third threshold setting unit determines whether or not the detected frequency of each pixel is equal to or higher than a predetermined fifth threshold, and determines that the frequency of each pixel is equal to or higher than a predetermined fifth threshold. If so, the third threshold is increased.
  • the condition for changing the motion vector of the target correction target pixel can be relaxed by increasing the third threshold value for comparison with the block matching difference value of the search divided block.
  • the video display device may further include a frequency detection unit that detects a frequency of each pixel of the current frame image, and the motion vector correction unit increases a frequency of each pixel detected by the frequency detection unit. It is preferable to further include a fourth threshold value setting unit that sets the second threshold value so that the value decreases as the value increases.
  • the frequency detection unit detects the frequency of each pixel of the current frame image. Then, the fourth threshold value setting unit sets the second threshold value so that the value decreases as the frequency of each detected pixel increases.
  • the pixel frequency when the pixel frequency is large, the number of pixels selected as the correction target can be increased, and the motion vector can be corrected more reliably.
  • the video display device may further include a frequency detection unit that detects a frequency of each pixel of the current frame image, and the correction target pixel selection unit has a predetermined frequency of each pixel detected by the frequency detection unit. It is preferable to reduce the first threshold when it is determined whether the frequency of each pixel is equal to or higher than a predetermined fifth threshold.
  • the frequency detection unit detects the frequency of each pixel of the current frame image. Then, the correction target pixel selection unit determines whether or not the detected frequency of each pixel is equal to or higher than a predetermined fifth threshold, and determines that the frequency of each pixel is equal to or higher than a predetermined fifth threshold. If this is the case, the first threshold value is reduced.
  • the pixel frequency when the pixel frequency is large, the number of pixels selected as the correction target can be increased, and the motion vector can be corrected more reliably.
  • the motion vector detection unit divides the frame image into a plurality of divided blocks including a plurality of pixels, detects the motion vector for each of the divided blocks, and detects the current frame.
  • the absolute value of the difference between the luminance value of each pixel in the target divided block of the image and the luminance value of each pixel of the reference divided block referred to when detecting the motion vector of the past frame image is calculated.
  • a total value of the pixel difference absolute value in the target divided block is calculated for each divided block as a block matching difference value of the target divided block, and the divided block includes a plurality of small blocks including a plurality of pixels,
  • the motion vector correction unit calculates the depth information of the pixel of interest among the pixels constituting the current frame image and the adjacent pixel adjacent to the pixel of interest.
  • a depth information difference value calculation unit that calculates the maximum absolute difference value with respect to the raw depth information for each pixel, and a maximum difference absolute value of the depth information of each pixel calculated by the depth information difference value calculation unit is Correction that corrects the motion vector for the small block including pixels that are equal to or larger than a predetermined first threshold and whose block matching difference value calculated by the motion vector detection unit is equal to or larger than a predetermined second threshold. From the correction target block selection unit to be selected as the target block and the target correction target block among the correction target blocks selected by the correction target block selection unit in a direction in which the maximum value of the absolute difference value of the depth information decreases.
  • a predetermined search range is determined for each correction target block in which the block matching difference value is equal to or smaller than a predetermined third threshold value. And when the block search unit finds a divided block that is equal to or smaller than the third threshold, the block search unit finds a motion vector of each pixel included in the target correction target block. And a motion vector changing unit that changes the motion vector of the divided block.
  • the motion vector detection unit divides the frame image into a plurality of divided blocks including a plurality of pixels, and a motion vector is detected for each of the divided blocks, and the current frame image in the target divided block is detected.
  • the absolute value of the difference between the luminance value of each pixel and the luminance value of each pixel of the reference divided block referenced when detecting the motion vector of the past frame image is calculated.
  • a total value in the target divided block is calculated for each divided block as a block matching difference value of the target divided block.
  • the divided block includes a plurality of small blocks including a plurality of pixels.
  • the depth information difference value calculation unit calculates, for each pixel, the maximum difference absolute value between the depth information of the target pixel of the pixels constituting the current frame image and the depth information of the adjacent pixel adjacent to the target pixel. The Then, the maximum value of the difference absolute value of the depth information of each pixel calculated by the correction target block selection unit is equal to or greater than a predetermined first threshold, and the block matching difference value calculated by the motion vector detection unit is a predetermined value. A small block including pixels that are equal to or greater than the second threshold is selected as a correction target block for correcting a motion vector.
  • the block matching difference value is equal to or smaller than a predetermined third threshold value in a direction in which the maximum value of the difference absolute value of the depth information decreases from the target correction target block among the correction target blocks selected by the block search unit.
  • a certain divided block is searched within a predetermined search range for each correction target block.
  • the motion vector changing unit changes the motion vector of each pixel included in the target correction target block to the motion vector of the found divided block.
  • a small block including a pixel having a large maximum difference absolute value of the depth information of each pixel included in a divided block with low motion vector reliability is selected as a correction target block, the foreground image and the background image Can be reliably selected as a correction target block. Further, since the motion vector of the target correction target block is changed to the motion vector of the search divided block, a motion vector with low reliability can be corrected to a motion vector with high reliability.
  • the video signal includes a left-eye video signal created to be viewed with a left eye and a right-eye video signal created to be viewed with a right eye, and the depth
  • the information acquisition unit detects the depth information based on the left-eye video signal and the right-eye video signal.
  • the depth information since the depth information is detected based on the left-eye video signal and the right-eye video signal used for stereoscopic display, the depth information can be created even when the depth information is not input. it can.
  • the depth information acquisition unit detects the depth information based on the saturation of each pixel constituting the current frame image.
  • the depth information is detected based on the saturation of each pixel constituting the current frame image, the depth information can be created even when the depth information is not input.
  • the video display method detects a motion vector from the plurality of frame images in a video signal including a plurality of temporally continuous frame images, and uses the detected motion vector to detect the video signal.
  • a motion vector detection step of detecting a motion vector of a current frame image using at least two or more frame images that are temporally mixed, and displaying the current vector A depth information acquisition step for acquiring depth information indicating a position in the depth direction of each pixel constituting the frame image, and the motion vector detection step based on the depth information acquired in the depth information acquisition step.
  • a motion vector correction step for correcting the motion vector.
  • the motion vector of the current frame image is detected using at least two or more frame images that are temporally changed.
  • depth information acquisition step depth information representing the position in the depth direction of each pixel constituting the current frame image is acquired.
  • the motion vector correction step the detected motion vector is corrected based on the acquired depth information.
  • the foreground image and the background image are clearly distinguished by the depth information representing the position in the depth direction, and the motion vector of the pixel at the boundary between the foreground image and the background image is corrected, so that the motion vector is detected with high accuracy. It is possible to prevent image quality deterioration.
  • the video processing device detects a motion vector from the plurality of frame images in a video signal including a plurality of temporally continuous frame images, and uses the detected motion vector to detect the video signal.
  • a video processing apparatus that performs predetermined image processing on a motion vector detection unit that detects a motion vector of a current frame image using at least two or more frame images that are temporally mixed, and the current frame image
  • a depth information acquisition unit that acquires depth information indicating a position in the depth direction of each pixel that constitutes the image, and the motion detected by the motion vector detection unit based on the depth information acquired by the depth information acquisition unit
  • a motion vector correction unit for correcting the vector is provided for correcting the vector.
  • the motion vector detection unit detects the motion vector of the current frame image using at least two or more frame images that are temporally changed.
  • the depth information acquisition unit acquires depth information indicating the position in the depth direction of each pixel constituting the current frame image. Then, the detected motion vector is corrected by the motion vector correction unit based on the acquired depth information.
  • the foreground image and the background image are clearly distinguished by the depth information representing the position in the depth direction, and the motion vector of the pixel at the boundary between the foreground image and the background image is corrected, so that the motion vector is detected with high accuracy. It is possible to prevent image quality deterioration.
  • the video display device, the video display method, and the video processing device according to the present invention can detect a motion vector with high accuracy and can prevent deterioration of the image quality, from a plurality of temporally continuous frame images.
  • the present invention is useful as a video display device, a video display method, and a video processing device that detect a motion vector from the plurality of frame images in a video signal and perform a predetermined image process on the video signal using the detected motion vector. is there.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)
  • Control Of Indicators Other Than Cathode Ray Tubes (AREA)

Abstract

La présente invention se rapporte à un dispositif d'affichage vidéo, à un procédé d'affichage vidéo et à un dispositif de traitement vidéo qui permettent de détecter des vecteurs de mouvement avec un grand degré de précision et également d'empêcher la détérioration de la qualité d'image. Un dispositif d'affichage vidéo (10) est pourvu : d'une unité de détection de vecteur de mouvement (3) destinée à détecter un vecteur de mouvement d'une image actuelle à l'aide d'au moins deux images se succédant dans le temps ; d'une unité d'acquisition d'informations de profondeur (2) destinée à acquérir des informations de profondeur qui représentent les positions dans le sens de la profondeur des pixels qui composent l'image actuelle ; et d'une unité de correction de vecteur de mouvement (4) destinée à corriger le vecteur de mouvement détecté par l'unité de détection de vecteur de mouvement (3) sur la base des informations de profondeur acquises par l'unité d'acquisition d'informations de profondeur (2).
PCT/JP2012/000786 2011-04-25 2012-02-06 Dispositif d'affichage vidéo, procédé d'affichage vidéo et dispositif de traitement vidéo WO2012147247A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2011-097431 2011-04-25
JP2011097431 2011-04-25

Publications (1)

Publication Number Publication Date
WO2012147247A1 true WO2012147247A1 (fr) 2012-11-01

Family

ID=47071783

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2012/000786 WO2012147247A1 (fr) 2011-04-25 2012-02-06 Dispositif d'affichage vidéo, procédé d'affichage vidéo et dispositif de traitement vidéo

Country Status (1)

Country Link
WO (1) WO2012147247A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2017011397A (ja) * 2015-06-18 2017-01-12 キヤノン株式会社 画像処理装置、画像処理方法、及びプログラム
CN113792771A (zh) * 2021-08-31 2021-12-14 青岛信芯微电子科技股份有限公司 用于处理图卡的tcon芯片

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2002123842A (ja) * 2000-10-13 2002-04-26 Takumi:Kk 立体視画像生成装置および情報記憶媒体
JP3729933B2 (ja) * 1996-05-31 2005-12-21 松下電器産業株式会社 自動監視装置

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3729933B2 (ja) * 1996-05-31 2005-12-21 松下電器産業株式会社 自動監視装置
JP2002123842A (ja) * 2000-10-13 2002-04-26 Takumi:Kk 立体視画像生成装置および情報記憶媒体

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2017011397A (ja) * 2015-06-18 2017-01-12 キヤノン株式会社 画像処理装置、画像処理方法、及びプログラム
CN113792771A (zh) * 2021-08-31 2021-12-14 青岛信芯微电子科技股份有限公司 用于处理图卡的tcon芯片
CN113792771B (zh) * 2021-08-31 2024-03-08 青岛信芯微电子科技股份有限公司 用于处理图卡的tcon芯片

Similar Documents

Publication Publication Date Title
US8446524B2 (en) Apparatus and method for frame rate conversion
US8184703B2 (en) Interpolated frame generating method and interpolated frame generating apparatus
EP1804488A2 (fr) Estimateur de mouvement et procédé de mouvement
US20140098089A1 (en) Image processing device, image processing method, and program
US20070230830A1 (en) Apparatus for creating interpolation frame
US10462337B2 (en) Liquid crystal display device and image processing method for same
US20080069221A1 (en) Apparatus, method, and computer program product for detecting motion vector and for creating interpolation frame
JP5257248B2 (ja) 画像処理装置および方法、ならびに画像表示装置
US9161011B2 (en) Image processing apparatus and control method thereof
US8830257B2 (en) Image displaying apparatus
JP2012079251A (ja) 画像処理装置及び画像処理システム
US10368048B2 (en) Method for the representation of a three-dimensional scene on an auto-stereoscopic monitor
WO2012147247A1 (fr) Dispositif d'affichage vidéo, procédé d'affichage vidéo et dispositif de traitement vidéo
US20120162528A1 (en) Video processing device and video display device
JP2009055340A (ja) 画像表示装置及び方法、画像処理装置及び方法
JP5127973B1 (ja) 映像処理装置、映像処理方法および映像表示装置
JPH11231832A (ja) 動ベクトル検出方法、動画像表示方法及び動画像表示装置
US20110273449A1 (en) Video processing apparatus and video display apparatus
US20100214488A1 (en) Image signal processing device
JP4892105B1 (ja) 映像処理装置、映像処理方法および映像表示装置
EP1691545B1 (fr) Dispositif d'interpolation des lignes de balayage
JP2009181067A (ja) 画像表示装置及び方法、画像処理装置及び方法
JP5343159B1 (ja) 画像処理装置、画像処理方法および画像処理プログラム
US20110109795A1 (en) Image processing apparatus and image processing method
KR102279495B1 (ko) 디스플레이장치 및 그 제어 방법

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 12776568

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12776568

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP