Static identification area detection method and module, chip, electronic device and medium

Abstract

A static identification area detection method and module, chip, electronic device and medium, the method includes: taking each pixel point in the image frame to be detected as the center, determining the corresponding first image block, and taking the same-position pixel point in the next image frame as the center, determining the corresponding same-position image block; searching for the first similar image block and the second similar image block similar to the first image block in the previous image frame and the next image frame, and searching for the third similar image block similar to the corresponding same-position image block in the image frame to be detected; respectively calculating the corresponding first motion vector, second motion vector and third motion vector; based on the corresponding first, second and third similar image blocks and the corresponding first, second and third motion vectors, judging whether the corresponding pixel point in the image frame to be detected is a pixel point in the static identification area. The above scheme can improve the accuracy of static identification area detection.

Description

Technical Field

[0001] The present invention relates to the field of image recognition technology, and in particular to a static identification area detection method and module, chip, electronic equipment and medium. Background Art

[0002] Generally, in a personal computer (PC) or a high definition television (HDTV), frame rate conversion (FRC) may be performed for compatibility between programs having various broadcast signal standards such as the PAL system or the NTSC system.

[0003] Frame rate conversion refers to the conversion of the number of frames output per second, and its purpose is to solve the problem of poor visual effects when playing moving pictures. In particular, when the frame rate is increased, it may be necessary to insert new frames, called insert frames. In order to generate the insert frames, motion estimation (Motion Estimation, ME) can be used. Motion estimation can include searching for the most similar blocks between the previous frame and the current frame. Motion vector (Motion Vector, MV) can represent the amplitude of block motion during motion estimation.

[0004] In the frame rate conversion process, it is generally necessary to separate the static mark area in the image. However, the existing static mark area detection method of the video image has the problem of low detection accuracy. Summary of the invention

[0005] The problem solved by the present invention is to provide a static identification area detection method to improve the accuracy of static identification area detection.

[0006] In order to solve the above problems, the present invention provides a static identification area detection method, the method comprising:

[0007] Taking each pixel point in the image frame to be detected as the center, respectively, a corresponding first image block is determined in the image frame to be detected, and taking the same-position pixel point in the next image frame of the image frame to be detected as the center, a corresponding same-position image block is determined in the next image frame; the coordinate position of the same-position pixel point in the next image frame is the same as the coordinate position of the pixel point in the image frame to be detected;

[0008] Searching for an image block similar to the corresponding first image block in the previous image frame and the next image frame of the image frame to be detected, and searching for an image block similar to the corresponding co-located image block in the image frame to be detected, as the corresponding first similar image block, the second similar image block and the third similar image block respectively;

[0009] respectively calculating a first motion vector from the corresponding first image block to the corresponding first similar image block and a second motion vector from the corresponding first image block to the corresponding second similar image block, and calculating a third motion vector from the corresponding co-located image block to the corresponding third similar image block;

[0010] Based on the corresponding first similar image block, second similar image block and third similar image block and the corresponding first motion vector, second motion vector and third motion vector, it is determined whether the corresponding pixel point in the image frame to be detected is a pixel point in the static identification area.

[0011] Optionally, after determining whether the similarity calculation of the corresponding first similar image block, the second similar image block and the third similar image block and the motion vector calculation of the first motion vector, the second motion vector and the third motion vector are reliable, and whether the directions of the corresponding first motion vector, the second motion vector and the third motion vector are consistent, the method further includes:

[0012] Based on the information of similarity calculation reliability of the corresponding first similar image block, the second similar image block and the third similar image block and motion vector calculation reliability of the corresponding first motion vector, the second motion vector and the third motion vector, and the information of consistency of motion vector directions of the corresponding first motion vector, the second motion vector and the third motion vector, a corresponding local confidence identifier is set for the corresponding pixel point in the image frame to be detected;

[0013] Taking the corresponding pixel point as the center, determining a corresponding second image block in the image frame to be detected;

[0014] Based on the local confidence identifiers of all pixels in the corresponding second image block, calculate the local confidence information of the corresponding pixel relative to the corresponding second image block; the local confidence information is used to identify the consistency of the corresponding pixel with other pixels in the corresponding second image in the time domain and the spatial domain;

[0015] Based on the calculated local confidence information of the corresponding pixel point relative to the corresponding second image block, it is determined whether the pixel point in the image frame to be detected is a pixel point in the static identification area.

[0016] Accordingly, an embodiment of the present invention further provides a static identification area detection module, including:

[0017] The determining unit is adapted to determine the corresponding first image block in the image frame to be detected with each pixel point in the image frame to be detected as the center, and determine the corresponding co-located image block in the next image frame of the image frame to be detected with the co-located pixel point in the next image frame as the center; the coordinate position of the co-located pixel point in the next image frame is the same as the coordinate position of the pixel point in the image frame to be detected;

[0018] A searching unit, adapted to search for an image block similar to the corresponding first image block in the previous image frame and the next image frame, and to search for an image block similar to the corresponding co-located image block in the image frame to be detected, as the corresponding first similar image block, second similar image block and third similar image block respectively;

[0019] a calculation unit, adapted to respectively calculate a first motion vector from the corresponding first image block to the corresponding first similar image block and a second motion vector from the corresponding first image block to the corresponding second similar image block, and calculate a third motion vector from the corresponding co-located image block to the corresponding third similar image block;

[0020] The detection unit is adapted to determine whether the corresponding pixel point in the image frame to be detected is a pixel point in the static identification area based on the corresponding first similar image block, the second similar image block and the third similar image block and the corresponding first motion vector, the second motion vector and the third motion vector.

[0021] Correspondingly, an embodiment of the present invention further provides a chip on which the above-mentioned detection module is integrated.

[0022] Correspondingly, an embodiment of the present invention also provides a device, comprising at least one memory and at least one processor, wherein the memory stores one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement any of the detection methods described above.

[0023] Correspondingly, an embodiment of the present invention further provides a storage medium, wherein the storage medium stores one or more computer instructions, and the one or more computer instructions are used to implement any of the detection methods described above.

[0024] Compared with the prior art, the technical solution of the present invention has the following advantages:

[0025] The static identification area detection method in the embodiment of the present invention is as follows: taking each pixel point in the image frame to be detected as the center, determining the corresponding first image block in the image frame to be detected, and taking the co-located pixel point in the next image frame of the image frame to be detected as the center, determining the corresponding co-located image block in the next image frame; searching for image blocks similar to the corresponding first image block in the previous image frame and the next image frame, and searching for image blocks similar to the corresponding co-located image block in the image frame to be detected, respectively as the corresponding first similar image block, the second similar image block and the third similar image block; respectively calculating the first motion vector from the corresponding first image block to the corresponding first similar image block and the second motion vector from the corresponding first image block to the corresponding second similar image block, and calculating the third motion vector from the corresponding co-located image block to the corresponding third similar image block; based on the corresponding first similar image block, the second similar image block and the third similar image block and the corresponding first motion vector, the second motion vector and the third motion vector, judging whether the corresponding pixel point in the image frame to be detected is a pixel point in the static identification area. Because the corresponding first similar image block, second similar image block and third similar image block and the corresponding first motion vector, second motion vector and third motion vector can be used to determine the consistency information of the corresponding pixel points in the image frame to be detected in the time domain and spatial domain of the previous and next frames of the video, the accuracy of static identification area recognition can be improved.

[0026] Further, after determining whether the similarity calculation of the corresponding first similar image block, the second similar image block and the third similar image block is reliable, or whether the calculation of the corresponding first motion vector, the second motion vector and the third motion vector is reliable, or whether the directions of the corresponding first motion vector, the second motion vector and the third motion vector are consistent, the method further includes: based on the information of the reliability of the similarity calculation of the corresponding first similar image block, the second similar image block and the third similar image block and the reliability of the motion vector calculation of the corresponding first motion vector, the second motion vector and the third motion vector, and the motion vectors of the corresponding first motion vector, the second motion vector and the third motion vector The information of consistency of direction is used to set a corresponding local confidence mark for the corresponding pixel point in the image frame to be detected; the corresponding second image block is determined in the image frame to be detected with the corresponding pixel point as the center; based on the local confidence marks of all pixels in the corresponding second image block, the local confidence information of the corresponding pixel point relative to the corresponding second image block is calculated; the local confidence information is used to identify the consistency information of the corresponding pixel point with other pixels in the corresponding second image in the time domain and the spatial domain; based on the calculated local confidence information of the corresponding pixel point relative to the corresponding second image block, it is determined whether the pixel point in the image frame to be detected is a pixel point in the static identification area. Because the local confidence information can be used to identify the consistency information of the corresponding pixel point with other pixels in the corresponding second image in the time domain and the spatial domain, the accuracy of static identification area recognition can be further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 A schematic diagram of a flow chart of a static identification area detection method in an embodiment of the present invention;

[0028] Figure 2 A schematic diagram of a flow chart of another static identification area detection method in an embodiment of the present invention;

[0029] Figure 3 Schematic diagram of a flow chart of another static identification area detection method in an embodiment of the present invention;

[0030] Figure 4 It is a schematic diagram of calculating the local execution degree of similarity of a current pixel point with respect to a first image block in an embodiment of the present invention;

[0031] Figure 5 Schematic diagram of the framework structure of a static identification area detection module in an embodiment of the present invention;

[0032] Figure 6 A hardware structure diagram of an electronic device provided in accordance with an embodiment of the present invention. DETAILED DESCRIPTION

[0033] As can be seen from the background technology, the existing static marking area detection method has the problem of low detection accuracy.

[0034] According to the different motion states, objects in video image frames can generally be divided into static objects and moving objects. Static objects are objects in a static state in the video image frame, such as background objects, text subtitles, etc.; moving objects are objects in a moving state in the video image frame, such as moving vehicles, running people, etc.

[0035] Static logos are a type of static objects in a video image frame. They include logos that are static for a short or long time, such as station logos, subtitles, or menu graphical interfaces (On-Screen Display, OSD), and are at the top layer of the image in the presented video image.

[0036] However, the existing static sign area detection method cannot accurately distinguish the background object from the static sign area, and has the problem of low recognition accuracy.

[0037] To solve the above problems, the present invention provides a static identification area detection method, comprising: taking each pixel point in an image frame to be detected as the center, determining a corresponding first image block in the image frame to be detected, and taking the co-located pixel point in the next image frame of the image frame to be detected as the center, determining a corresponding co-located image block in the next image frame; searching for an image block similar to the corresponding first image block in the previous image frame and the next image frame, and searching for an image block similar to the corresponding co-located image block in the image frame to be detected, as the corresponding first similar image block, the second similar image block and the third similar image block, respectively; calculating a first motion vector from the corresponding first image block to the corresponding first similar image block and a second motion vector from the corresponding first image block to the corresponding second similar image block, and calculating a third motion vector from the corresponding co-located image block to the corresponding third similar image block; judging whether the corresponding pixel point in the image frame to be detected is a pixel point in the static identification area based on the corresponding first similar image block, the second similar image block and the third similar image block and the corresponding first motion vector, the second motion vector and the third motion vector.

[0038] In the static identification area detection method provided by an embodiment of the present invention, the corresponding first similar image block, second similar image block and third similar image block and the corresponding first motion vector, second motion vector and third motion vector can be used to determine the consistency information of the corresponding pixel points in the image frame to be detected in the time domain and spatial domain of the previous and next frames of the video, so as to identify the static identification area in the video image, which can improve the accuracy of static identification.

[0039] Figure 1 FIG. 1 is a flow chart showing a static identification area detection method in an embodiment of the present invention. Figure 1 A static identification area detection method is used to detect a static identification area in a video image, which may specifically include:

[0040] Step S101, taking each pixel point in the image frame to be detected as the center, determining a corresponding first image block in the image frame to be detected, and taking a co-located pixel point in the next image frame of the image frame to be detected as the center, determining a corresponding co-located image block in the next image frame; the coordinate position of the co-located pixel point in the next image frame is the same as the coordinate position of the pixel point in the image frame to be detected;

[0041] Step S102, searching for an image block similar to the corresponding first image block in the previous image frame and the next image frame of the image frame to be detected, and searching for an image block similar to the corresponding co-located image block in the image frame to be detected, as the corresponding first similar image block, the second similar image block and the third similar image block respectively;

[0042] Step S103, respectively calculating a first motion vector from the corresponding first image block to the corresponding first similar image block and a second motion vector from the corresponding first image block to the corresponding second similar image block, and calculating a third motion vector from the corresponding co-located image block to the corresponding third similar image block;

[0043] Step S104 , based on the corresponding first similar image block, second similar image block and third similar image block and the corresponding first motion vector, second motion vector and third motion vector, determining whether the corresponding pixel point in the image frame to be detected is a pixel point in the static identification area.

[0044] The above scheme, based on the first similar image block, the second similar image block and the third similar image block and the corresponding first motion vector, the second motion vector and the third motion vector, can determine the consistency information of the corresponding pixel points in the image frame to be detected in the time domain and the spatial domain, and can improve the accuracy of static identification area recognition.

[0045] In order to make the above-mentioned purposes, features and advantages of the embodiments of the present invention more obvious and easy to understand, the following is a summary of the present invention. Figures 2 to 4 The specific embodiments of the present invention are described in detail.

[0046] See also Figure 2 A static identification area detection method in an embodiment of the present invention may specifically include the following steps:

[0047] Step S201, obtaining an image frame to be detected, and traversing the pixels in the image frame to be detected to obtain the current pixel point traversed to.

[0048] For the convenience of description, the image frame to be detected is recorded as f(t), and the current pixel point traversed in the image frame to be detected is recorded as f(x0, y0, t). Among them, t represents the time corresponding to the image frame to be detected, and x0 and y0 represent the horizontal axis (or X-axis) coordinate and the vertical axis (or Y-axis) coordinate of the current pixel point to be traversed, respectively.

[0049] Step S202, taking the current pixel point traversed as the center, determining the corresponding first image block in the image frame to be detected, and taking the same-position pixel point in the next image frame of the image frame to be detected as the center, determining the corresponding same-position image block in the next image frame.

[0050] The previous image frame of the image frame to be detected f(t) is recorded as f(t-1), and the next image frame of the image frame to be detected f(t) is recorded as f(t+1).

[0051] The co-located pixel point is a pixel point in the next image frame f(t+1), and its coordinate position is the same as the coordinate position of the current pixel point f(x0, y0, t) to be detected in the image frame f(t) to be detected, and is recorded as f(x0, y0, (t+1)).

[0052] The first image block is an image block of size M*N (M and N are both integers greater than 1) centered at the current pixel point f(x0, y0, t) traversed in the current image frame to be detected f(t), denoted as Block1(t). The size of the first image block Block1(t), that is, the values ​​of M and N, can be set according to the requirements of computational complexity and accuracy, as long as the size of the set first image block Block1(t) is smaller than the size of the image frame to be detected, and is not limited here.

[0053] The co-located image block is an image block centered at the co-located pixel point f(x0, y0, (t+1)) in the next image frame f(t+1), and is recorded as Block1(t+1). The corresponding co-located image block Block1(t+1) has the same size as the corresponding first image block Block1(t).

[0054] Step S203, searching for an image block similar to the first image block in the previous image frame and the next image frame of the image frame to be detected, and searching for an image block similar to the co-located image block in the image frame to be detected, as the first similar image block, the second similar image block and the third similar image block respectively.

[0055] In a specific implementation, an existing search algorithm, such as a global algorithm, a three-step search algorithm or a diamond search algorithm, etc., can be used to search for image blocks similar to the first image block Block1(t) in the previous image frame f(t-1) and the next image frame f(t+1), respectively, and search for image blocks similar to the co-located image block Block1(t+1) in the image frame to be detected f(t), and the first similar image block, the second similar image block and the third similar image block obtained by the search are respectively recorded as SB1(t-1), SB2(t+1) and SB3(t).

[0056] In one embodiment of the present invention, the corresponding similar image blocks are determined by calculating the Sum of Absolute Differences (SAD). Specifically, the first similar image block SB1(t-1) is an image block with the minimum SAD with the first image block Block1(t) in the previous image frame f(t-1), the second similar image block SB2(t+1) is an image block with the minimum SAD with the first image block Block1(t) in the next image frame f(t+1), and the third similar image block SB3(t) is an image block with the minimum SAD with the co-located image block Block1(t+1) in the image frame to be detected f(t).

[0057] In one embodiment of the present invention, the following formula is used to calculate the sum of absolute errors between corresponding image blocks:

[0058]

[0059] Among them, D(i,j) represents the absolute error sum between corresponding image blocks, (i,j) is the starting point position located at the upper left corner of the corresponding image block, S(i+M,j+N) represents the corresponding image block, T(M,N) represents the corresponding matching template, W represents the width of the corresponding image block, and H represents the height of the corresponding image block.

[0060] When the value of D(i,j) calculated by the above formula (1) is the smallest, it is determined that the M*N pixel block starting from the position (i,j) is the best match for the matching template T(M,N), and the image block corresponding to D(i,j) is taken as the corresponding similar image block.

[0061] It is understandable that other methods may be used to determine the first similar image block SB1(t-1), the second similar image block SB2(t+1) and the third similar image block SB3(t). Those skilled in the art may make a selection based on actual needs and no limitation is made here.

[0062] Step S204 , respectively calculating a first motion vector from the first image block to the first similar image block and a second motion vector from the first image block to the second similar image block, and calculating a third motion vector from the co-located image block to the third similar image block.

[0063] The first motion vector, the second motion vector and the third motion vector are respectively denoted as vec1(x0, y0), vec2(x0, y0) and vec3(x0, y0). The first motion vector vec1(x0, y0) is used to indicate the information of the motion direction between the first similar image block SB1(t-1) and the first image block Block1(t); the second motion vector vec2(x0, y0) is used to indicate the information of the motion direction between the first image block Block1(t) and the second similar image block SB2(t+1); the third motion vector vec3(x0, y0) is used to indicate the information of the motion direction between the co-located image block Block1(t+1) and the third similar image block SB3(t).

[0064] In a specific implementation, the calculation method of the first motion vector vec1 (x0, y0), the second motion vector vec2 (x0, y0) and the third motion vector vec3 (x0, y0) can be performed using the existing motion vector calculation algorithm between image blocks, which is not limited here.

[0065] Step S205, determining whether the similarity calculation of the first similar image block, the second similar image block and the third similar image block is reliable, or whether the motion vector calculation of the first motion vector, the second motion vector and the third motion vector is reliable, or whether the directions of the first motion vector, the second motion vector and the third motion vector are consistent; when the determination result is yes, step S206 can be executed; otherwise, step S207 can be executed.

[0066] In a specific implementation, when the first similar image block SB1(t-1), the second similar image block SB2(t+1) and the third similar image block SB3(t) are determined by calculating the minimum sum of absolute errors (SAD), it is determined whether the similarity calculation of the first similar image block SB1(t-1), the second similar image block SB2(t+1) and the third similar image block SB3(t) is reliable, that is, it is determined whether the sum of absolute errors corresponding to the first similar image block SB1(t-1), the second similar image block SB2(t+1) and the third similar image block SB3(t) is reliable.

[0067] In one embodiment of the present invention, a method for determining whether the absolute errors and calculations corresponding to the first similar image block SB1(t-1), the second similar image block SB2(t+1) and the third similar image block SB3(t) are reliable includes:

[0068] Obtain the absolute error sum SAD1 between the first similar graphic block SB1(t-1) and the first image block Block1(t), the absolute error sum SAD2 between the second similar graphic block SB2(t+1) and the first image block Block1(t), and the absolute error sum SAD3 between the third similar graphic block SB3(t) and the co-located image block Block1(t+1).

[0069] The absolute error sum SAD1 between the first similar graphic block SB1(t-1) and the first image block Block1(t) is compared with a preset first absolute error sum threshold sad_th1; when it is determined that the absolute error sum SAD1 between the first similar graphic block SB1(t-1) and the first image block Block1(t) is less than the first absolute error sum threshold sad_th1, it is determined that the absolute error sum calculation of the first similar image block SB1(t-1) is reliable; otherwise, it is determined that the absolute error sum calculation of the first similar image block SB1(t-1) is not reliable.

[0070] The absolute error sum SAD2 between the second similar graphic block SB2(t+1) and the first image block Block1(t) is compared with the second absolute error sum threshold sad_th2. When it is determined that the absolute error sum SAD2 between the second similar graphic block SB2(t+1) and the first image block Block1(t) is less than the second absolute error sum threshold sad_th2, it is determined that the absolute error sum calculation of the second similar graphic block SB2(t+1) is reliable; otherwise, it is determined that the absolute error sum calculation of the second similar graphic block SB2(t+1) is not reliable.

[0071] The absolute error sum SAD3 between the third similar graphic block SB3(t) and the co-located image block Block1(t+1) is compared with the third absolute error sum threshold sad_th3. When it is determined that the absolute error sum SAD3 between the third similar graphic block SB3(t) and the co-located image block Block1(t+1) is less than the third absolute error sum threshold sad_th3, it is determined that the absolute error sum calculation of the third similar graphic block SB3(t) is reliable; otherwise, it is determined that the absolute error sum calculation of the third similar graphic block SB3(t) is not reliable.

[0072] In a specific implementation, the first absolute error and threshold sad_th1, the second absolute error and threshold sad_th2, and the third absolute error and threshold sad_th3 can be set to be the same or different according to actual needs, which is not limited here.

[0073] It can be understood that when the absolute errors and calculations corresponding to the first similar image block SB1(t-1), the second similar image block SB2(t+1, and the third similar image block SB3(t) are all reliable, it indicates that the similarity between the current pixel point on the previous and next frames of the video image is large, while the similarity between the pixel points in the static identification area on the previous and next frames of the video image is small. Therefore, when the absolute errors and calculations of the first similar image block SB1(t-1), the second similar image block SB2(t+1, and the third similar image block SB3(t) are reliable, it indicates that the current pixel point has the possibility of being a pixel point in the non-static identification area.

[0074] In one embodiment of the present invention, a method for determining whether motion vector calculation of a first motion vector vec1(x0, y0), a second motion vector vec2(x0, y0) and a third motion vector vec3(x0, y0) is reliable includes:

[0075] The horizontal component vec1(x0) of the first motion vector vec1(x0, y0) is compared with the first horizontal component threshold vecx th1, and the vertical component vec1(y0) of the first motion vector vec1(x0, y0) is compared with the first vertical component threshold vecyth1; when the horizontal component vec1(x0) of the first motion vector vec1(x0, y0) is greater than the first horizontal component threshold vecx th1, and the vertical component vec1(y0) of the first motion vector vec1(x0, y0) is greater than the first vertical component threshold vecyth1, it is determined that the calculation of the corresponding first motion vector vec1(x0, y0) is reliable; otherwise, the calculation of the corresponding first motion vector vec1(x0, y0) is not reliable;

[0076] The horizontal component vec2(x0) of the second motion vector vec2(x0, y0) is compared with the second horizontal component threshold vecx th2, and the vertical component vec2(y0) of the second motion vector vec2(x0, y0) is compared with the second vertical component threshold vecyth2; when the horizontal component vec2(x0) of the second motion vector vec2(x0, y0) is greater than the second horizontal component threshold vecx th2, and the vertical component vec2(y0) of the second motion vector vec2(x0, y0) is greater than the second vertical component threshold vecyth2, the calculation of the corresponding second motion vector vec2(x0, y0) is reliable; otherwise, the calculation of the corresponding second motion vector vec2(x0, y0) is not reliable;

[0077] The horizontal component vec3(x0) of the third motion vector vec3(x0, y0) is compared with the third horizontal component threshold vecx th3, and the vertical component vec3(y0) of the third motion vector vec3(x0, y0) is compared with the third vertical component threshold vecyth3; when the horizontal component vec3(x0) of the third motion vector vec3(x0, y0) is greater than the third horizontal component threshold vecx th3, and the vertical component vec3(y0) of the third motion vector vec3(x0, y0) is greater than the third vertical component threshold vecyth3, the calculation of the corresponding third motion vector vec3(x0, y0) is reliable; otherwise, the calculation of the corresponding third motion vector vec3(x0, y0) is not reliable;

[0078] In a specific implementation, the first horizontal component threshold vecx th1, the second horizontal component threshold vecx th2 and the third horizontal component threshold vecx th3 can be set to be the same or different according to actual needs; the first vertical component threshold vecy th1, the second vertical component threshold vecy th2 and the third vertical component threshold vecy th3 can be set to be the same or different according to actual needs, and are not limited here.

[0079] It can be understood that when the motion vector calculation of the first motion vector vec1 (x0, y0), the second motion vector vec2 (x0, y0) and the third motion vector vec3 (x0, y0) is reliable, it indicates that the motion vector of the current pixel point f (x0, y0, t) in the direction of the front and back frames of the video is large, and the pixel point in the static identification area belongs to a static object, and its motion vector is a motion vector of approximately zero. Therefore, when the motion vector calculation of the first motion vector vec1 (x0, y0), the second motion vector vec2 (x0, y0) and the third motion vector vec3 (x0, y0) is reliable, it indicates that the pixel point has the possibility of being a pixel point in the non-static identification area.

[0080] In one embodiment of the present invention, a method for determining whether the directions of a first motion vector vec1(x0, y0) and a second motion vector vec2(x0, y0) and a third motion vector vec2(x0, y0) are consistent includes:

[0081] Calculate the absolute sum of the difference between the horizontal component vec1(x0) of the first motion vector vec1(x0, y0) and the horizontal component vec2(x0) of the second motion vector vec2(x0, y0) |vec1(x0)-vec2(x0)|, ​​and calculate the absolute sum of the difference between the vertical component vec1(y0) of the first motion vector vec1(x0, y0) and the vertical component vec2(y0) of the second motion vector vec1(x0, y0) |vec1(y0)-vec2(y0)|, and calculate the sum of |vec1(x0)-vec2(x0)| and |vec1(y0)-vec2(y0)| as the first absolute difference sum dif_1;

[0082] Calculate the absolute sum of the difference between the horizontal component vec1(x0) of the first motion vector vec1(x0, y0) and the horizontal component vec3(x0) of the third motion vector vec3(x0, y0) |vec1(x0)-vec3(x0)|, ​​calculate the absolute sum of the difference between the vertical component vec1(y0) of the first motion vector vec1(x0, y0) and the vertical component vec3(y0) of the third motion vector vec1(x0, y0) |vec1(y0)-vec3(y0)|, and calculate the sum of |vec1(x0)-vec3(x0)| and |vec1(y0)-vec3(y0)| as the second absolute difference sum dif_2;

[0083] When it is determined that the sum of the first absolute difference dif_1 and the second absolute difference dif_2 is less than the absolute difference threshold dif_th, it is determined that the directions of the first motion vector vec1 (x0, y0) and the second motion vector vec2 (x0, y0) and the third motion vector vec3 (x0, y0) are consistent.

[0084] In a specific implementation, the absolute difference and the threshold dif_th can be set according to actual needs and are not limited here.

[0085] It can be understood that when the direction of the first motion vector vec1 (x0, y0) is consistent with the direction of the second motion vector vec2 (x0, y0) and the third motion vector vec3 (x0, y0), it indicates that the motion vector of the current pixel is consistent in the direction of the front and back frames of the video image, while the motion vector of the pixel in the static identification area is less consistent in the direction of the front and back frames of the video image. Therefore, when the motion vector calculation of the first motion vector vec1 (x0, y0), the second motion vector vec2 (x0, y0) and the third motion vector vec3 (x0, y0) is reliable, it indicates that the current pixel has the possibility of being a pixel in the non-static identification area.

[0086] Step S206: determine the current pixel point traversed to be a pixel point in the non-static identification area.

[0087] In a specific implementation, when it is determined that the similarity calculation of the first similar image block SB1(t-1), the second similar image block SB2(t+1) and the third similar image block SB3(t) is reliable, or the motion vector calculation of the first motion vector vec1(x0, y0), the second motion vector vec2(x0, y0) and the third motion vector vec3(x0, y0) is reliable, or the directions of the first motion vector vec1(x0, y0) and the second motion vector vec2(x0, y0) and the third motion vector vec3(x0, y0) are consistent, the current pixel point f(x0, y0, t) traversed can be identified as a pixel point within the non-static identification area.

[0088] Step S207 is executed to determine the current pixel point traversed to be a pixel point in the static identification area.

[0089] In a specific implementation, when it is determined that the similarity calculation of the first similar image block SB1(t-1), the second similar image block SB2(t+1) and the third similar image block SB3(t) is not reliable, and the motion vector calculation of the first motion vector vec1(x0, y0), the second motion vector vec2(x0, y0) and the third motion vector vec3(x0, y0) is not reliable, and the directions of the first motion vector vec1(x0, y0) and the second motion vector vec2(x0, y0) and the third motion vector vec3(x0, y0) are not consistent, the current pixel point f(x0, y0, t) traversed is identified as a pixel point within the non-static identification area.

[0090] Execute step S208 to determine whether all pixels in the image frame to be detected have been traversed; if the determination result is yes, the subsequent steps can be executed. Figure 3 Otherwise, step S209 can be executed.

[0091] Execute step S209 to obtain the next pixel point in the image frame to be detected, use the obtained next pixel point as the current pixel point to be traversed, and start executing from step S202 until all pixel points in the image frame to be detected are traversed.

[0092] By executing the above steps S201 to S209, it is possible to determine whether the pixel is a pixel in the static identification area by the characteristics of each pixel in the image frame to be detected in terms of similarity, motion vector and consistency of motion vector direction on the previous and next frames of the video. In another embodiment of the present invention, in order to further improve the accuracy of detection, a corresponding similarity local confidence mark can be set for the corresponding pixel based on whether the similarity calculation of the corresponding first similar image block, the second similar image block and the third similar image block is reliable; a corresponding motion vector local confidence mark is set for the corresponding pixel based on whether the motion vector calculation of the corresponding first motion vector, the second motion vector and the third motion vector is reliable; and a corresponding motion vector consistency local confidence mark is set for the corresponding pixel based on whether the direction between the first motion vector and the second motion vector and the third motion vector is consistent.

[0093] Specifically:

[0094] When the similarity calculation of the corresponding first similar image block SB1(t-1), the second similar image block SB2(t+1) and the third similar image block SB3(t) is reliable, the similarity local confidence flag sad_flag_mux(x, y, t) of the corresponding pixel point (x, y, t) is set to the first value; conversely, when the similarity calculation of the corresponding first similar image block SB1(t-1), the second similar image block SB2(t+1) and the third similar image block SB3(t) is not reliable, the similarity local confidence flag sad_flag_mux(x, y, t) of the corresponding pixel point (x, y, t) is set to the second value.

[0095] When the motion vector calculation of the corresponding first motion vector vec1(x0, y0), the second motion vector vec2(x0, y0) and the third motion vector vec3(x0, y0) is reliable, the motion vector local confidence flag vec_flag_mux(x, y, t) of the corresponding pixel point (x, y, t) is set to the third value; conversely, when the motion vector calculation of the corresponding first motion vector vec1(x0, y0), the second motion vector vec2(x0, y0) and the third motion vector vec3(x0, y0) is not reliable, the corresponding similarity local confidence flag is set to the fourth value.

[0096] When the direction of the corresponding first motion vector vec1 (x0, y0) is consistent with the second motion vector vec2 (x0, y0) and the third motion vector vec3 (x0, y0), the motion vector consistency local confidence flag vec_flag_mux (x, y, t) of the corresponding pixel point (x, y, t) is set to the fifth value; conversely, when the direction of the corresponding first motion vector vec1 (x0, y0) is inconsistent with the second motion vector vec2 (x0, y0) and the third motion vector vec3 (x0, y0), the corresponding motion vector consistency local confidence flag is set to the sixth value.

[0097] In a specific implementation, the first value, the second value, the third value, the fourth value, the fifth value and the sixth value can be set according to actual needs, as long as the first value and the second value are different, the third value and the fourth value are different, and the fifth value and the sixth value are different. In one embodiment of the present invention, the first value, the third value and the fifth value are the same, all of which are 1; the second value, the fourth value and the sixth value are the same, all of which are 0.

[0098] By adopting the above method, after setting the corresponding similarity local confidence flag, motion vector local confidence flag and motion vector consistency local confidence flag for each pixel in the image frame to be detected, the similarity local confidence flag, motion vector local confidence flag and motion vector consistency local confidence flag corresponding to the pixel can be used to compare the consistency of the pixel with other surrounding pixels in the time domain and space domain to further determine whether the pixel is a pixel in the static identification area. For details, please refer to Figure 3 :

[0099] Step S301 is executed to traverse all pixel points in the image frame to be detected and obtain the current pixel point traversed to.

[0100] Step S302 is executed to determine a corresponding second image block in the image frame to be detected, taking the current pixel point traversed as the center.

[0101] The second image block includes the pixel point to be detected currently in the image frame to be detected, and other pixel points located within the area of ​​the pixel point to be detected currently, which is recorded as Block2(t).

[0102] In a specific implementation, the size of the second image block Block2(t) is a*b. Those skilled in the art can understand that the size of the second image block Block2(t) can be set to a*b according to actual needs, where a is the number of pixels in each row of the second image block Block2(t), b is the number of pixels in each column of the second image block Block2(t), and both a and b are integers greater than or equal to 2.

[0103] In a specific implementation, the second image block Block2(t) may be the same as or different from the first image block Block1(t), that is, the value of a may be the same as or different from the value of M, and the value of b may be the same as or different from the value of N. Technicians in this field may make settings according to actual needs and no limitation is made here.

[0104] Step S303 is executed to calculate the local confidence of similarity, the local confidence of the motion vector and the local confidence of the consistency of the motion vector of the current pixel relative to the corresponding second image block.

[0105] In one embodiment of the present invention, a method for calculating a local confidence doc1 of a similarity between a current pixel point f(x0, y0, t) and a corresponding second image block Block2(t) includes:

[0106] First, obtain the similarity local confidence indicator sad_flag_mux(x, y) of each pixel point f(x, y, t) in the second image block Block2(t) corresponding to the current pixel point f(x0, y0, t). The calculation method of the similarity local confidence indicator sad_flag_mux(x, y) of each pixel point f(x, y, t) in the second image block Block2(t) can be found in the introduction of the corresponding part above, which will not be repeated here.

[0107] See also Figure 4 , when the similarity local confidence identifier sad_flag_mux(x, y, t) of each pixel point f(x, y, t) in the corresponding second image block Block2(t) is obtained, the similarity local confidence identifier sad_flag_mux(x, y, t) of each pixel point f(x, y, t) is placed at the corresponding pixel point position (x, y) respectively, and the corresponding first identification matrix sad_flag_mux(t) can be obtained.

[0108] Afterwards, the first identification matrix sad_flag_mux(t) is multiplied and then summed with the preset first identification template mask1(t) to obtain the local confidence doc1 of the similarity of the current pixel point f(x0, y0, t) to the second image block Block2(t). The first identification template mask1(t) and the first identification matrix sad_flag_mux(t) are matrices of the same size, that is, both are matrices of size a*b, and each element in the first identification template mask1(t) is a non-negative integer.

[0109] The method for calculating the local confidence of the motion vector of the current pixel point f(x0, y0, t) relative to the corresponding second image block Block2(t) and the local confidence of the motion vector consistency is similar to the above-mentioned method for calculating the local confidence of the motion vector of the current pixel point f(x0, y0, t) relative to the second image block Block2(t). Specifically:

[0110] In one embodiment of the present invention, a method for calculating a local confidence doc2 of a motion vector of a current pixel point f(x0, y0, t) relative to a corresponding second image block Block2(t) includes:

[0111] Get the motion vector local confidence flag vec_flag_mux(x, y, t) corresponding to each pixel point f(x, y, t) in the corresponding second image block Block2(t). The calculation method of the motion vector local confidence flag vec_flag_mux(x, y) of each pixel point f(x, y, t) in the second image block Block2(t) is as described in the introduction of the corresponding part above, which will not be repeated here.

[0112] When the motion vector local confidence identifier vec_flag_mux(x, y, t) corresponding to each pixel point f(x, y, t) in the corresponding second image block Block2(t) is obtained, the motion vector local confidence identifier vec_flag_mux(x, y, t) corresponding to each pixel point f(x, y, t) in the corresponding second image block Block2(t) is placed at the corresponding pixel point position (x, y) respectively, and the corresponding second identification matrix vec_flag_mux(t) can be obtained.

[0113] Finally, the second identification matrix vec_flag_mux(t) is multiplied and then summed to obtain the local confidence doc2 of the motion vector of the current pixel point f(x0, y0, t) relative to the second image block Block2(t). The second identification template mask2(t) and the second identification matrix vec_flag_mux(t) are matrices of the same size, that is, both are matrices of size a*b, and each element in the second identification template mask2(t) is a non-negative integer.

[0114] In one embodiment of the present invention, a method for calculating a local confidence doc3 of a motion vector consistency of a current pixel point f(x0, y0, t) relative to the second image block Block2(t) includes:

[0115] First, obtain the motion vector consistency local confidence indicator dif_flag_mux(x, y, t) corresponding to each pixel point f(x, y, t) in the corresponding second image block Block2(t). The calculation method of the motion vector consistency local confidence indicator dif_flag_mux(x, y, t) of each pixel point f(x, y, t) in the second image block Block2(t) is as described in the introduction of the corresponding part above, which will not be repeated here.

[0116] Afterwards, the motion vector consistency local confidence flag dif_flag_mux(x, y, t) corresponding to each pixel point f(x, y, t) in the acquired second image block Block1(t) is placed at the corresponding pixel point position (x, y), and the corresponding third flag matrix dif_flag_mux(t) can be obtained.

[0117] Finally, the third identification matrix dif_flag_mux(t) is multiplied and then summed to obtain the local confidence doc3 of the motion vector consistency of the pixel point f(x0, y0, t) relative to the second image block Block2(t). The third identification template mask3(t) and the third identification matrix dif_flag_mux(t) are matrices of the same size, that is, both are matrices of a*b size, and each element in the third identification template mask3(t) is a non-negative integer.

[0118] Step S304, respectively calculate the first contribution score, the second contribution score and the third contribution score corresponding to the local confidence of similarity, the local confidence of the motion vector and the local confidence of the consistency of the motion vector of the current pixel relative to the corresponding second image block.

[0119] In one embodiment of the present invention, the first contribution score score1, the second contribution score score2 and the third contribution score score3 corresponding to the local confidence doc1 of the similarity, the local confidence doc2 of the motion vector and the local confidence doc3 of the motion vector consistency of the current pixel point f(x0, y0, t) can be calculated in the following manner:

[0120] (a) When calculating the first contribution score score1 corresponding to the similarity local confidence doc1, the similarity local confidence doc1 of the current pixel point f(x0, y0, t) is compared with the preset first similarity local confidence threshold sad_flag_th1 and the second similarity local confidence threshold sad_flag_th2 respectively; when the similarity local confidence doc1 of the current pixel point f(x0, y0, t) is less than the first similarity local confidence threshold sad_flag_th1, the corresponding first contribution score score1 is set is the seventh value; when the similarity local confidence doc1 of the current pixel point f(x0, y0, t) is greater than or equal to the first similarity local confidence threshold sad_flag_th1 and less than the second similarity local confidence threshold sad_flag_th2, the corresponding first contribution score score1 is set to the eighth value; when the similarity local confidence doc1 of the current pixel point f(x0, y0, t) is greater than or equal to the second similarity local confidence threshold sad_flag_th2, the corresponding first contribution score score1 is set to the ninth value.

[0121] The first similarity local confidence threshold sad_flag_th1 and the second similarity local confidence threshold sad_flag_th2 can be set according to actual needs, as long as the set first similarity local confidence threshold sad_flag_th1 is smaller than the second similarity local confidence threshold sad_flag_th2.

[0122] (b) When calculating the second contribution score score2 corresponding to the local confidence score doc2 of the motion vector of the current pixel point f(x0, y0, t), the local confidence score doc2 of the motion vector of the current pixel point f(x0, y0, t) is compared with the preset first local confidence score threshold vec_flag_th1 and the second local confidence score threshold vec_flag_th2 of the motion vector; when the local confidence score doc2 of the motion vector of the current pixel point f(x0, y0, t) is less than the first local confidence score threshold vec_flag_th1 of the motion vector, the corresponding second contribution score score2 is calculated. The second contribution score score2 is set to the tenth value; when the local confidence doc2 of the motion vector of the current pixel point f(x0, y0, t) is greater than or equal to the first motion vector local confidence threshold vec_flag_th1 and less than the second motion vector local confidence threshold vec_flag_th2, the corresponding second contribution score score2 is set to the eleventh value; when the corresponding motion vector local confidence doc2 is greater than or equal to the second motion vector local confidence threshold vec_flag_th2, the corresponding second contribution score score2 is set to the twelfth value.

[0123] The first motion vector local confidence threshold vec_flag_th1 and the second motion vector local confidence threshold vec_flag_th2 can be set according to actual needs, as long as the set first motion vector local confidence threshold vec_flag_th1 is smaller than the second motion vector local confidence threshold vec_flag_th2.

[0124] (c) When calculating the third contribution score score3 corresponding to the local confidence score doc3 of the motion vector consistency of the current pixel point f(x0, y0, t), the local confidence score doc3 of the motion vector consistency of the current pixel point f(x0, y0, t) is compared with the preset first local confidence score dif_flag_th1 and second local confidence score dif_flag_th2 of the motion vector consistency; when the local confidence score doc3 of the motion vector consistency of the current pixel point f(x0, y0, t) is less than the first local confidence score dif_flag_th1, the corresponding local confidence score score3 is calculated. The third contribution score score3 is set to the thirteenth value; when the motion vector consistency local confidence doc3 of the current pixel point f(x0, y0, t) is greater than or equal to the first motion vector consistency local confidence threshold dif_flag_th1 and less than the second motion vector consistency local confidence threshold dif_flag_th2, the corresponding third contribution score score3 is set to the fourteenth value; when the motion vector consistency local confidence doc3 is greater than or equal to the threshold second motion vector consistency local confidence threshold dif_flag_th2, the corresponding third contribution score score3 is set to the fifteenth value.

[0125] Among them, the first motion vector consistency local confidence threshold dif_flag_th1 and the second motion vector consistency local confidence threshold dif_flag_th2 can be set according to actual needs, as long as the set first motion vector consistency local confidence threshold dif_flag_th1 is smaller than the second motion vector consistency local confidence threshold dif_flag_th2.

[0126] It can be understood that the seventh to fifteenth values ​​can be set according to detection requirements, as long as the seventh value, eighth value and ninth value are different, the tenth value, eleventh value and twelfth value are different, and the thirteenth value, fourteenth value and fifteenth value are different, without any limitation here.

[0127] In one embodiment of the present invention, the seventh value, the tenth value and the thirteenth value are 0, the eighth value, the eleventh value and the fourteenth value are 1, and the ninth value, the twelfth value and the fifteenth value are all 2.

[0128] Execute step S305 to calculate the recognition score corresponding to the current pixel based on the calculated first contribution score, second contribution score, and third contribution score.

[0129] In one embodiment of the present invention, when calculating the first contribution score score1, the second contribution score score2 and the third contribution score score3 corresponding to the current pixel point, the recognition score Score corresponding to the pixel point is calculated using the following formula:

[0130] Score=(score1*a1)+(score2*a2)+(score3*a3) (1)

[0131] and:

[0132] a1+a2+a3=1 (2)

[0133] Among them, a1, a2 and a3 represent the weights corresponding to the first contribution score, the second contribution score and the third contribution score respectively.

[0134] Execute step S306 to determine whether the recognition score corresponding to the current pixel is greater than or equal to the preset recognition threshold; when the judgment result is yes, execute step S307; otherwise, execute step S308.

[0135] In a specific implementation, the recognition threshold score_th can be set according to the need of static identification area recognition, and is not limited here.

[0136] Execute step S307 to mark the current pixel as a pixel in the non-static marking area.

[0137] In a specific implementation, when it is determined that the recognition score score corresponding to the previous pixel point f(x0, y0, t) is greater than or equal to the preset recognition threshold score_th, it indicates that the current pixel point f(x0, y0, t) is consistent with other pixel points f(x, y, t) in the second image block Block2(t) in the time domain and the spatial domain, and meets the characteristics of the pixel points in the non-static identification area, so it can be determined that the current pixel point f(x0, y0, t) in the image frame to be detected is a pixel point in the non-static identification area. At this time, the logo_flag corresponding to the current pixel point f(x0, y0, t) can be set to a corresponding value, such as 1, so that the current pixel point f(x0, y0, t) can be determined as a pixel point in the non-static identification area in the future through the value of the logo_flag.

[0138] Execute step S308 to identify the current pixel as a pixel in the non-static identification area.

[0139] In a specific implementation, when it is determined that the recognition score score corresponding to the previous pixel point f(x0, y0, t) is less than the preset recognition threshold score_th, it indicates that the current pixel point f(x0, y0, t) is inconsistent with other pixel points f(x, y, t) in the second image block Block2(t) in both the time domain and the spatial domain, and meets the characteristics of the pixel points in the static identification area, so it can be determined that the current pixel point f(x0, y0, t) in the image frame to be detected is a pixel point in the static identification area. At this time, the logo_flag corresponding to the current pixel point f(x0, y0, t) can be set to a corresponding value, such as 0, so that the current pixel point f(x0, y0, t) can be determined as a pixel point in the static identification area in the future through the value of the logo_flag.

[0140] Execute step S309 to determine whether all the pixel points in the image frame to be detected have been traversed; when the judgment result is yes, end the operation; when the judgment result is no, you can execute step S310.

[0141] Execute step S310 to obtain the next pixel point in the image frame to be detected, use the next pixel point as the current pixel point to be traversed, and restart execution from step S302 until all pixel points in the image frame to be detected are traversed.

[0142] The above is a detailed description of the static marking area detection method in the embodiment of the present invention. The following will introduce the modules corresponding to the static marking area detection method.

[0143] Figure 5 FIG. 1 is a schematic diagram showing a framework structure of a static identification area detection module in an embodiment of the present invention. Figure 5 , a static identification area detection module 50 may include a determination unit 501, a search unit 502, a calculation unit 503 and a detection unit 504, wherein:

[0144] The determining unit 501 is adapted to determine the corresponding first image block in the image frame to be detected with each pixel point in the image frame to be detected as the center, and determine the corresponding co-located image block in the image frame next to the image frame to be detected with the co-located pixel point in the image frame next to the image frame to be detected as the center; the coordinate position of the co-located pixel point in the next image frame is the same as the coordinate position of the pixel point in the image frame to be detected;

[0145] The searching unit 502 is adapted to search for an image block similar to the corresponding first image block in the previous image frame and the next image frame of the image frame to be detected, and to search for an image block similar to the corresponding co-located image block in the image frame to be detected, as the corresponding first similar image block, the second similar image block and the third similar image block, respectively;

[0146] The calculation unit 503 is adapted to respectively calculate a first motion vector from the corresponding first image block to the corresponding first similar image block and a second motion vector from the corresponding first image block to the corresponding second similar image block, and calculate a third motion vector from the corresponding co-located image block to the corresponding third similar image block;

[0147] The detection unit 504 is adapted to determine whether the corresponding pixel point in the image frame to be detected is a pixel point within the static identification area based on the corresponding first similar image block, the second similar image block and the third similar image block and the corresponding first motion vector, the second motion vector and the third motion vector.

[0148] In one embodiment of the present invention, the search unit 502 is suitable for searching for an image block having a minimum sum of absolute errors with the corresponding first image block in the previous image frame and the next image frame of the image frame to be detected, and searching for an image block having a minimum sum of absolute errors with the corresponding co-located image block in the image frame to be detected, as the first similar image block, the second similar image block and the third similar image block.

[0149] In one embodiment of the present invention, the detection unit 504 is adapted to determine whether the similarity calculation of the corresponding first similar image block, the second similar image block and the third similar image block is reliable, or whether the motion vector calculation of the corresponding first motion vector, the second motion vector and the third motion vector is reliable, or whether the directions of the corresponding first motion vector, the second motion vector and the third motion vector are consistent; when it is determined that the similarity calculation of the corresponding first similar image block, the second similar image block and the third similar image block is reliable, or the motion vector calculation of the corresponding first motion vector, the second motion vector and the third motion vector is reliable, or the directions of the corresponding first motion vector, the second motion vector and the third motion vector are consistent, the corresponding pixel point is determined to be a pixel point in the non-static identification area; otherwise, the corresponding pixel point is determined to be a pixel point in the static identification area.

[0150] In one embodiment of the present invention, the detection unit 504 is adapted to determine that the similarity calculation of the corresponding first similar image block is reliable when the sum of absolute errors between the corresponding first similar graphic block and the corresponding first image block is less than a first absolute error and a threshold value; otherwise, determine that the similarity calculation of the corresponding first similar image block is not reliable; when the sum of absolute errors between the corresponding second similar graphic block and the corresponding first image block is less than a second absolute error and a threshold value, determine that the similarity calculation of the corresponding second similar image block is reliable; otherwise, determine that the similarity calculation of the corresponding second similar image block is not reliable; when the sum of absolute errors between the corresponding third similar graphic block and the corresponding co-located image block is less than a third absolute error and a threshold value, determine that the similarity calculation of the corresponding third similar image block is not reliable; otherwise, determine that the similarity calculation of the corresponding third similar image block is not reliable.

[0151] In one embodiment of the present invention, the detection unit 504 is suitable for determining that the calculation of the corresponding first motion vector is reliable when the horizontal component of the corresponding first motion vector is greater than the first horizontal component threshold, and the vertical component of the corresponding first motion vector is greater than the first vertical component threshold; otherwise, it is determined that the calculation of the corresponding first motion vector is not reliable; when the horizontal component of the corresponding second motion vector is greater than the second horizontal component threshold, and the vertical component of the corresponding second motion vector is greater than the second vertical component threshold, it is determined that the calculation of the corresponding second motion vector is reliable; otherwise, it is determined that the calculation of the corresponding second motion vector is not reliable; when the horizontal component of the corresponding third motion vector is greater than the third horizontal component threshold, and the vertical component of the corresponding third motion vector is greater than the third vertical component threshold, it is determined that the calculation of the corresponding third motion vector is reliable; otherwise, it is determined that the calculation of the corresponding third motion vector is not reliable.

[0152] In one embodiment of the present invention, the detection unit 504 is suitable for calculating the absolute sum of the differences between the horizontal components of the corresponding first motion vector and the corresponding second motion vector and the absolute sum of the differences between the vertical components of the corresponding first motion vector and the corresponding second motion vector as the first absolute difference sum; calculating the absolute sum of the differences between the horizontal components of the corresponding first motion vector and the corresponding third motion vector and the absolute sum of the differences between the vertical components of the corresponding first motion vector and the corresponding third motion vector as the second absolute difference sum; when it is determined that the sum of the first absolute difference sum and the second absolute difference sum is less than the absolute difference sum threshold, it is determined that the directions of the corresponding first motion vector, the second motion vector and the third motion vector are consistent; otherwise, it is determined that the directions of the corresponding first motion vector, the corresponding second motion vector and the corresponding third motion vector are inconsistent.

[0153] In one embodiment of the present invention, the detection unit 504 is further adapted to, after determining whether the similarity calculations of the corresponding first similar image block, the second similar image block and the third similar image block are similar, whether the motion vector calculations of the corresponding first motion vector, the second motion vector and the third motion vector are reliable, and whether the directions of the corresponding first motion vector, the second motion vector and the third motion vector are consistent, determine whether the similarity calculations of the corresponding first similar image block, the second similar image block and the third similar image block are similar, based on information about the reliability of the motion vector calculations of the corresponding first motion vector, the second motion vector and the third motion vector, and the motion vector calculation reliability of the corresponding first motion vector, the second motion vector and the third motion vector, and the motion vector calculation reliability of the corresponding first motion vector, the second motion vector and the third motion vector. The method comprises the following steps: according to the present invention, the method comprises the following steps: setting a corresponding local confidence mark for a pixel in the image frame to be detected based on the consistency of the motion vector direction; determining a corresponding second image block in the image frame to be detected with each pixel in the image frame to be detected as the center; calculating the local confidence information of the corresponding pixel relative to the corresponding second image block based on the local confidence mark of all pixels in the corresponding second image block; the local confidence information is used to identify the consistency information of the corresponding pixel with other pixels in the corresponding second image in the time domain and the spatial domain; and determining whether the pixel in the image frame to be detected is a pixel in the static identification area based on the calculated local confidence information of the corresponding pixel relative to the corresponding second image block.

[0154] In a specific implementation, the local confidence identifier includes at least one of a similarity local confidence identifier, a motion vector local confidence identifier, and a motion vector consistency local confidence identifier; the local confidence information of the corresponding pixel point relative to the corresponding second image block includes at least one of a similarity local confidence, a motion vector local confidence, and a motion vector consistency local confidence.

[0155] In one embodiment of the present invention, the detection unit 504 is adapted to set the similarity local confidence flag of the corresponding pixel point to a first value when it is determined that the similarity calculation of the corresponding first similar image block, the second similar image block and the third similar image block is reliable; otherwise, the similarity local confidence flag of the corresponding pixel point is set to a second value.

[0156] In one embodiment of the present invention, the detection unit 504 is suitable for setting the local confidence flag of the motion vector of the corresponding pixel point to a third value when it is determined that the motion vector calculation of the corresponding first motion vector, the second motion vector and the third motion vector is reliable; otherwise, the local confidence flag of the motion vector of the corresponding pixel point is set to a fourth value.

[0157] In one embodiment of the present invention, the detection unit 504 is suitable for setting the local confidence flag of the motion vector consistency of the corresponding pixel point to a fifth value when the directions of the corresponding first motion vector, second motion vector and third motion vector are consistent; otherwise, the local confidence flag of the motion vector consistency of the corresponding pixel point is set to a sixth value.

[0158] In one embodiment of the present invention, the detection unit 504 is suitable for calculating the local confidence of similarity, local confidence of motion vector and local confidence of motion vector consistency of the corresponding pixel point relative to the corresponding second image block based on the local confidence identifier of similarity, local confidence identifier of motion vector and local confidence identifier of motion vector consistency of all pixels in the corresponding second image block; calculating the first contribution score, the second contribution score and the third contribution score corresponding to the local confidence of similarity, the local confidence of motion vector and the local confidence of motion vector consistency respectively; calculating the recognition score of the corresponding pixel point based on the calculated first contribution score, the second contribution score and the third contribution score; when the calculated recognition score is less than the preset recognition threshold, the corresponding pixel point is identified as a pixel point in the non-static identification area.

[0159] In one embodiment of the present invention, the detection unit 504 is suitable for respectively obtaining the similarity local confidence identifiers of all pixels in the corresponding second image block, and placing the obtained similarity local confidence identifiers of all pixels in the corresponding second image block at the corresponding pixel positions to form a corresponding first identification matrix; performing a dot multiplication and then a sum operation on the corresponding first identification matrix and the first identification template to obtain the similarity local confidence of the pixel relative to the corresponding second image block.

[0160] In one embodiment of the present invention, the detection unit 504 is suitable for respectively obtaining the local confidence identifiers of the motion vectors of all pixels in the corresponding second image block, and placing the obtained local confidence identifiers of the motion vectors of all pixels in the corresponding second image block at the corresponding pixel positions to form a corresponding second identification matrix; performing a dot multiplication and then a sum operation on the corresponding second identification matrix and the second identification template to obtain the local confidence of the motion vector of the corresponding pixel relative to the corresponding second image block.

[0161] In one embodiment of the present invention, the detection unit 504 is suitable for respectively obtaining the local confidence identifiers of the consistency of the motion vectors of all the pixels in the corresponding second image block, and placing the obtained local confidence identifiers of the consistency of the motion vectors of all the pixels in the corresponding second image block at the corresponding pixel positions to form a corresponding third identification matrix; performing a dot multiplication and then a sum operation on the corresponding third identification matrix and the third identification template to obtain the local confidence identifiers of the consistency of the motion vectors of the corresponding pixels relative to the corresponding second image block.

[0162] In one embodiment of the present invention, the detection unit 504 is suitable for setting the corresponding first contribution score to a preset seventh value when the similarity local confidence of the corresponding pixel point is less than a preset first similarity local confidence threshold; setting the corresponding first contribution score to an eighth value when the similarity local confidence of the corresponding pixel point is greater than or equal to the first similarity local confidence threshold and less than a preset second similarity local confidence threshold; and setting the corresponding first contribution score to a ninth value when the similarity local confidence of the corresponding pixel point is greater than or equal to the second similarity local confidence threshold.

[0163] In one embodiment of the present invention, the detection unit 504 is suitable for setting the corresponding second contribution score to the tenth value when the local confidence of the motion vector of the corresponding pixel point is less than the preset first local confidence threshold of the motion vector; setting the corresponding second contribution score to the eleventh value when the local confidence of the motion vector of the corresponding pixel point is greater than or equal to the first local confidence threshold of the motion vector and less than the preset second local confidence threshold of the motion vector; and setting the corresponding second contribution score to the twelfth value when the local confidence of the motion vector of the corresponding pixel point is greater than or equal to the second local confidence threshold of the motion vector.

[0164] In one embodiment of the present invention, the detection unit 504 is suitable for setting the corresponding third contribution score to the thirteenth value when the local confidence of the motion vector consistency of the corresponding pixel point is less than the preset first motion vector consistency local confidence threshold; when the local confidence of the motion vector consistency of the corresponding pixel point is greater than or equal to the first motion vector consistency local confidence threshold and less than the preset second motion vector consistency local confidence threshold, setting the corresponding third contribution score to the fourteenth value; when the local confidence of the motion vector consistency of the corresponding pixel point is greater than or equal to the second motion vector consistency local confidence threshold, setting the corresponding third contribution score to the fifteenth value.

[0165] In one embodiment of the present invention, the detection unit 504 is adapted to obtain a weighted sum of the corresponding first contribution score, the second contribution score and the third contribution score as the recognition score of the corresponding pixel point.

[0166] Correspondingly, an embodiment of the present invention further provides a chip, on which the static marking area detection module is integrated. The static marking area detection module is described in the above part and will not be described in detail.

[0167] Correspondingly, an embodiment of the present invention further provides an electronic device, which can execute the above-mentioned static identification area detection method in the form of a loaded program.

[0168] An optional hardware structure of the electronic device provided by the embodiment of the present invention can be as follows Figure 6 As shown, it includes: at least one processor 01, at least one communication interface 02, at least one memory 03 and at least one communication bus 04.

[0169] In the embodiment of the present invention, the number of the processor 01 , the communication interface 02 , the memory 03 , and the communication bus 04 is at least one, and the processor 01 , the communication interface 02 , and the memory 03 communicate with each other through the communication bus 04 .

[0170] The communication interface 02 may be an interface of a communication module for network communication, such as an interface of a GSM module.

[0171] The processor 01 may be a central processing unit CPU, or an application specific integrated circuit ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present invention.

[0172] The memory 03 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0173] The memory 03 stores one or more computer instructions, and the one or more computer instructions are executed by the processor 01 to implement the auxiliary graphics configuration method of the embodiment of the present invention.

[0174] It should be noted that the electronic device implemented above may also include other devices (not shown) that may not be necessary for understanding the contents disclosed in the embodiments of the present invention; since these other devices may not be necessary for understanding the contents disclosed in the embodiments of the present invention, the embodiments of the present invention will not introduce them one by one.

[0175] An embodiment of the present invention further provides a storage medium, wherein the storage medium stores one or more computer instructions, and the one or more computer instructions are used to implement the static identification area detection method provided by the embodiment of the present invention.

[0176] The above-mentioned embodiments of the present invention are combinations of elements and features of the present invention. Unless otherwise mentioned, the elements or features may be considered as optional. Each element or feature may be put into practice without being combined with other elements or features. In addition, the embodiments of the present invention may be constructed by combining some elements and / or features. The order of operations described in the embodiments of the present invention may be rearranged. Some constructions of any one embodiment may be included in another embodiment, and may be replaced by the corresponding construction of another embodiment. It is obvious to those skilled in the art that claims that do not have a clear reference relationship to each other in the attached claims may be combined into embodiments of the present invention, or may be included as new claims in the amendment after submitting this application.

[0177] The embodiments of the present invention can be implemented by various means such as hardware, firmware, software or a combination thereof. In a hardware configuration, the method according to the exemplary embodiment of the present invention can be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, etc.

[0178] In a firmware or software configuration, the embodiments of the present invention may be implemented in the form of modules, procedures, functions, etc. The software code may be stored in a memory unit and executed by a processor. The memory unit is located inside or outside the processor and may send data to and receive data from the processor via various known means.

[0179] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0180] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the scope defined by the claims.

Claims

1. A static marking area detection method for detecting a static marking area in a video image, characterized in that: include: Taking each pixel point in the image frame to be detected as the center, respectively, a corresponding first image block is determined in the image frame to be detected, and taking the same-position pixel point in the next image frame of the image frame to be detected as the center, a corresponding same-position image block is determined in the next image frame; the coordinate position of the same-position pixel point in the next image frame is the same as the coordinate position of the pixel point in the image frame to be detected; Searching for an image block similar to the corresponding first image block in the previous image frame and the next image frame of the image frame to be detected, and searching for an image block similar to the corresponding co-located image block in the image frame to be detected, as the corresponding first similar image block, second similar image block and third similar image block respectively; respectively calculating a first motion vector from the corresponding first image block to the corresponding first similar image block and a second motion vector from the corresponding first image block to the corresponding second similar image block, and calculating a third motion vector from the corresponding co-located image block to the corresponding third similar image block; Based on the corresponding first similar image block, second similar image block and third similar image block and the corresponding first motion vector, second motion vector and third motion vector, the consistency information of the corresponding pixel point in the image frame to be detected in the time domain and spatial domain of the previous and next frames of the video is determined, so as to determine whether the corresponding pixel point in the image frame to be detected is a pixel point within the static identification area.

2. The static marking area detection method according to claim 1, characterized in that: The step of searching for an image block similar to the corresponding first image block in a previous image frame and a next image frame of the image frame to be detected, and searching for an image block similar to the corresponding co-located image block in the image frame to be detected, comprises: An image block having a minimum sum of absolute errors with the corresponding first image block is searched in the previous image frame and the next image frame respectively, and an image block having a minimum sum of absolute errors with the corresponding co-located image block is searched in the image frame to be detected, which are respectively used as the first similar image block, the second similar image block and the third similar image block.

3. The static marking area detection method according to claim 2, characterized in that: The determining, based on the corresponding first similar image block, the second similar image block, and the third similar image block and the corresponding first motion vector, the second motion vector, and the third motion vector, whether the corresponding pixel point in the image frame to be detected is a pixel point in the static identification area includes: Determining whether the similarity calculation of the corresponding first similar image block, the second similar image block and the third similar image block is reliable, or whether the motion vector calculation of the corresponding first motion vector, the second motion vector and the third motion vector is reliable, or whether the directions of the corresponding first motion vector, the second motion vector and the third motion vector are consistent; When it is determined that the similarity calculation of the corresponding first similar image block, the second similar image block and the third similar image block is reliable, or the motion vector calculation of the corresponding first motion vector, the second motion vector and the third motion vector is reliable, or the directions of the corresponding first motion vector, the second motion vector and the third motion vector are consistent, the corresponding pixel point is determined to be a pixel point within the non-static identification area; otherwise, the corresponding pixel point is determined to be a pixel point within the static identification area.

4. The static marking area detection method according to claim 3, characterized in that: The determining whether the similarity calculation of the corresponding first similar image block, the second similar image block and the third similar image block is reliable includes: When the sum of the absolute errors between the corresponding first similar graphic block and the corresponding first image block is less than the first absolute error and threshold, determining that the similarity calculation of the corresponding first similar image block is reliable; Otherwise, it is determined that the similarity calculation of the corresponding first similar image block is not reliable; When the sum of the absolute errors between the corresponding second similar graphic block and the corresponding first image block is less than the second absolute error and threshold, determining that the similarity calculation of the corresponding second similar image block is reliable; Otherwise, it is determined that the similarity calculation of the corresponding second similar image block is not reliable; When the sum of the absolute errors between the corresponding third similar graphic block and the corresponding co-located image block is less than the third absolute error and threshold, determining that the similarity calculation of the corresponding third similar image block is reliable; Otherwise, it is determined that the similarity calculation of the corresponding third similar image block is not reliable.

5. The static marking area detection method according to claim 3, characterized in that: The determining whether the motion vector calculation of the first motion vector, the second motion vector, and the third motion vector is reliable includes: When the horizontal component of the corresponding first motion vector is greater than the first horizontal component threshold, and the vertical component of the corresponding first motion vector is greater than the first vertical component threshold, it is determined that the calculation of the corresponding first motion vector is reliable; otherwise, it is determined that the calculation of the corresponding first motion vector is not reliable; When the horizontal component of the corresponding second motion vector is greater than the second horizontal component threshold, and the vertical component of the corresponding second motion vector is greater than the second vertical component threshold, it is determined that the calculation of the corresponding second motion vector is reliable; otherwise, it is determined that the calculation of the corresponding second motion vector is not reliable; When the horizontal component of the corresponding third motion vector is greater than the third horizontal component threshold, and the vertical component of the corresponding third motion vector is greater than the third vertical component threshold, the calculation of the corresponding third motion vector is determined to be reliable; otherwise, the calculation of the corresponding third motion vector is determined to be unreliable.

6. The static marking area detection method according to claim 3, characterized in that: The determining whether the directions of the corresponding first motion vector, the second motion vector and the third motion vector are consistent includes: Calculate the sum of the absolute sum of the differences between the horizontal components of the corresponding first motion vector and the corresponding second motion vector and the absolute sum of the differences between the vertical components of the corresponding first motion vector and the corresponding second motion vector as a first absolute difference sum; calculating the sum of the absolute sum of the differences between the horizontal components of the corresponding first motion vector and the corresponding third motion vector and the absolute sum of the differences between the vertical components of the corresponding first motion vector and the corresponding third motion vector as a second absolute difference sum; When it is determined that the sum of the first absolute difference sum and the second absolute difference sum is less than the absolute difference sum threshold, it is determined that the corresponding first motion vector is consistent with the direction of the second motion vector and the third motion vector; otherwise, it is determined that the corresponding first motion vector is inconsistent with the direction of the corresponding second motion vector and the corresponding third motion vector.

7. The static marking area detection method according to claim 3, characterized in that: After determining whether the similarity calculations of the corresponding first similar image block, the second similar image block, and the third similar image block are similar, whether the motion vector calculations of the corresponding first motion vector, the second motion vector, and the third motion vector are reliable, and whether the directions of the corresponding first motion vector, the second motion vector, and the third motion vector are consistent, the method further includes: Based on the information of similarity calculation reliability of the corresponding first similar image block, the second similar image block and the third similar image block and motion vector calculation reliability of the corresponding first motion vector, the second motion vector and the third motion vector, and the information of consistency of motion vector directions of the corresponding first motion vector, the second motion vector and the third motion vector, setting a corresponding local confidence flag for the pixel point in the image frame to be detected; Taking each pixel in the image frame to be detected as the center, determining a corresponding second image block in the image frame to be detected; Based on the local confidence identifiers of all pixels in the corresponding second image block, calculate the local confidence information of the corresponding pixel relative to the corresponding second image block; the local confidence information is used to identify the consistency of the corresponding pixel with other pixels in the corresponding second image block in the time domain and the spatial domain; Based on the calculated local confidence information of the corresponding pixel point relative to the corresponding second image block, it is determined whether the pixel point in the image frame to be detected is a pixel point in the static identification area.

8. The static marking area detection method according to claim 7, characterized in that: The local confidence mark includes at least one of a similarity local confidence mark, a motion vector local confidence mark and a motion vector consistency local confidence mark; The local confidence information of the corresponding pixel point relative to the corresponding second image block includes at least one of a similarity local confidence, a motion vector local confidence, and a motion vector consistency local confidence.

9. The static marking area detection method according to claim 8, characterized in that: Setting a corresponding similarity local confidence mark for a corresponding pixel point in the image frame to be detected includes: When it is determined that the similarity calculation of the corresponding first similar image block, the second similar image block and the third similar image block is reliable, the similarity local confidence flag of the corresponding pixel point is set to the first value; otherwise, the similarity local confidence flag of the corresponding pixel point is set to the second value.

10. The static marking area detection method according to claim 8, characterized in that: Setting a corresponding motion vector local confidence flag for a corresponding pixel point in the image frame to be detected includes: When it is determined that the motion vector calculation of the corresponding first motion vector, second motion vector and third motion vector is reliable, the local confidence flag of the motion vector of the corresponding pixel point is set to the third value; otherwise, the local confidence flag of the motion vector of the corresponding pixel point is set to the fourth value.

11. The static marking area detection method according to claim 8, characterized in that: Setting a corresponding motion vector consistency local confidence flag for a corresponding pixel point in the image frame to be detected includes: When the directions of the corresponding first motion vector, second motion vector and third motion vector are consistent, the local confidence flag of the motion vector consistency of the corresponding pixel point is set to the fifth value; otherwise, the local confidence flag of the motion vector consistency of the corresponding pixel point is set to the sixth value.

12. The static marking area detection method according to claim 8, characterized in that: The determining, based on the calculated local confidence information of the pixel point relative to the corresponding second image block, whether the pixel point in the image frame to be detected is a pixel point in the static identification area includes: Calculating the similarity local confidence, motion vector local confidence and motion vector consistency local confidence of the corresponding pixel relative to the corresponding second image block based on the similarity local confidence identifiers, motion vector local confidence identifiers and motion vector consistency local confidence identifiers of all pixels in the corresponding second image block respectively; Respectively calculating the first contribution score, the second contribution score and the third contribution score corresponding to the local confidence of the similarity, the local confidence of the motion vector and the local confidence of the consistency of the motion vector; Based on the calculated first contribution score, second contribution score, and third contribution score, calculate the recognition score of the corresponding pixel point; When the calculated recognition score is less than a preset recognition threshold, the corresponding pixel point is identified as a pixel point within the static identification area.

13. The static marking area detection method according to claim 12, characterized in that: The calculating the local confidence of the similarity of the corresponding pixel point with respect to the corresponding second image block includes: Respectively obtaining similarity local confidence marks of all pixels in the corresponding second image block, and placing the obtained similarity local confidence marks of all pixels in the corresponding second image block at corresponding pixel positions to form a corresponding first mark matrix; A dot multiplication is performed on the corresponding first identification matrix and the first identification template and then a sum operation is performed to obtain a local confidence level of similarity between the pixel point and the corresponding second image block.

14. The static marking area detection method according to claim 12, characterized in that: The calculating the local confidence of the motion vector of the corresponding pixel point relative to the corresponding second image block includes: Respectively obtain the local confidence identifiers of the motion vectors of all the pixels in the corresponding second image block, and place the obtained local confidence identifiers of the motion vectors of all the pixels in the corresponding second image block at the corresponding pixel positions to form a corresponding second identifier matrix; The corresponding second identification matrix is ​​multiplied with the second identification template and then summed to obtain a local confidence of a motion vector of the corresponding pixel point relative to the corresponding second image block.

15. The static marking area detection method according to claim 12, characterized in that: The calculating the local confidence of the consistency of the motion vector of the corresponding pixel point relative to the corresponding second image block includes: The local confidence identifiers of the consistency of the motion vectors of all the pixels in the corresponding second image block are respectively obtained, and the local confidence identifiers of the consistency of the motion vectors of all the pixels in the corresponding second image block are respectively placed at the corresponding pixel positions to form a corresponding third identification matrix; the corresponding third identification matrix is ​​dotted with the third identification template and then summed to obtain the local confidence of the consistency of the motion vector of the corresponding pixel relative to the corresponding second image block.

16. The static marking area detection method according to claim 12, characterized in that: The calculating the first contribution score corresponding to the local confidence of the similarity includes: When the similarity local confidence of the corresponding pixel point is less than a preset first similarity local confidence threshold, setting the corresponding first contribution score to a preset seventh value; When the similarity local confidence of the corresponding pixel point is greater than or equal to the first similarity local confidence threshold and less than a preset second similarity local confidence threshold, setting the corresponding first contribution score to an eighth value; When the similarity local confidence of the corresponding pixel point is greater than or equal to the second similarity local confidence threshold, the corresponding first contribution score is set to a ninth value.

17. The static marking area detection method according to claim 12, characterized in that: The calculating a second contribution score corresponding to the local confidence of the motion vector includes: When the local confidence of the motion vector of the corresponding pixel point is less than the preset first local confidence threshold of the motion vector, the corresponding second contribution score is set to the tenth value; When the local confidence of the motion vector of the corresponding pixel point is greater than or equal to the first local confidence threshold of the motion vector and less than the preset second local confidence threshold of the motion vector, setting the corresponding second contribution score to an eleventh value; When the local confidence of the motion vector of the corresponding pixel point is greater than or equal to the second local confidence threshold of the motion vector, the corresponding second contribution score is set to the twelfth value.

18. The static marking area detection method according to claim 12, characterized in that: The calculating a third contribution score corresponding to the local confidence level of the motion vector consistency includes: When the local confidence level of the motion vector consistency of the corresponding pixel point is less than the preset first motion vector consistency local confidence level threshold, setting the corresponding third contribution score to the thirteenth value; When the local confidence level of the motion vector consistency of the corresponding pixel point is greater than or equal to the first local confidence level of the motion vector consistency and less than the preset second local confidence level of the motion vector consistency, setting the corresponding third contribution score to the fourteenth value; When the local confidence of the motion vector consistency of the corresponding pixel point is greater than or equal to the second motion vector consistency local confidence threshold, the corresponding third contribution score is set to the fifteenth value.

19. The static marking area detection method according to claim 12, characterized in that: Based on the calculated first contribution score, the second contribution score, and the third contribution score, the recognition score of the corresponding pixel point is calculated, including: A weighted sum of the corresponding first contribution score, the second contribution score, and the third contribution score is calculated as the recognition score of the corresponding pixel point.

20. A static marking area detection module, used to detect a static marking area in a video image, characterized in that: include: The determining unit is adapted to determine the corresponding first image block in the image frame to be detected with each pixel point in the image frame to be detected as the center, and determine the corresponding co-located image block in the next image frame of the image frame to be detected with the co-located pixel point in the next image frame as the center; the coordinate position of the co-located pixel point in the next image frame is the same as the coordinate position of the pixel point in the image frame to be detected; A searching unit, adapted to search for an image block similar to the corresponding first image block in a previous image frame and a next image frame of the image frame to be detected, and to search for an image block similar to the corresponding co-located image block in the image frame to be detected, as the corresponding first similar image block, second similar image block and third similar image block, respectively; a calculation unit, adapted to respectively calculate a first motion vector from the corresponding first image block to the corresponding first similar image block and a second motion vector from the corresponding first image block to the corresponding second similar image block, and calculate a third motion vector from the corresponding co-located image block to the corresponding third similar image block; The detection unit is suitable for determining the consistency information of the corresponding pixel points in the image frame to be detected in the time domain and the spatial domain of the previous and next frames of the video based on the corresponding first similar image block, the second similar image block and the third similar image block and the corresponding first motion vector, the second motion vector and the third motion vector, so as to judge whether the corresponding pixel points in the image frame to be detected are the pixel points in the static identification area.

21. A chip, characterized in that: The chip is integrated with the detection module as claimed in claim 20.

22. An electronic device, characterized in that: It comprises at least one memory and at least one processor, wherein the memory stores one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the detection method according to any one of claims 1-19.

23. A storage medium, characterized in that: The storage medium stores one or more computer instructions, and the one or more computer instructions are used to implement the detection method according to any one of claims 1-19.

Images