WO2018082332A1

WO2018082332A1 - Image processing method and device, and robot

Info

Publication number: WO2018082332A1
Application number: PCT/CN2017/092029
Authority: WO
Inventors: 刘若鹏; 欧阳一村
Original assignee: 深圳光启合众科技有限公司
Priority date: 2016-11-07
Filing date: 2017-07-06
Publication date: 2018-05-11
Also published as: CN108090924B; CN108090924A

Abstract

Disclosed are an image processing method and device, and a robot. The method comprises: obtaining a color reference value of a target object in a target image frame, wherein the color reference value is a color value of the color of the target object that occurs most frequently in a preset area; obtaining a first area formed by pixels in the target image frame where the difference between the pixels and the color reference value is less than or equal to a preset value; and using the edge of the first area as the contour of the target object. The present application resolves the technical problem in the prior art of being unable to determine the contour of a target object in a video.

Description

Image processing method and device, robot

Technical field

The present application relates to the field of images, and in particular to an image processing method and apparatus, and a robot.

Background technique

The monitoring and tracking of target objects is an important part of the research of intelligent video systems. It plays an important role in the fields of technology, military, transportation, security, medical care, etc. In the prior art, the image frames of video are extracted and processed. The results obtained are of low precision and cannot accurately determine the contour of the target object in the video.

In response to the above problems, no effective solution has been proposed yet.

Summary of the invention

The embodiment of the present application provides an image processing method and apparatus, and a robot, to at least solve the technical problem that the contour of the target object in the video cannot be accurately determined in the prior art.

According to an aspect of an embodiment of the present application, an image processing method includes: acquiring a color reference value of a target object in a target image frame, wherein the color reference value is a number of occurrences of the target object in a preset area. a color value of a maximum color; acquiring a first region formed by a pixel point in the target image frame that has a difference from the color reference value that is less than or equal to a preset value; using the edge of the first region as the The outline of the target object.

Further, acquiring the color reference value of the target object in the target image frame includes: selecting a second region in which the target object is located from the target image frame; and acquiring the pre-predetermined from a center point of the second region A plurality of pixel points in the area are set; a color value corresponding to the color having the most occurrences is searched for from the plurality of pixel points; and a color value corresponding to the color having the most occurrences is used as the color reference value.

Further, the edge of the first region as the contour of the target object includes: acquiring a matrix formed by color values of all pixel points in the first region, wherein each element in the matrix represents one pixel a color value of the point; calculating a gradient of the matrix, and calculating an absolute value of the gradient; and mapping an absolute value of the gradient to the target image frame to obtain an outline of the target object.

Further, selecting the second region where the target object is located from the target image frame includes: acquiring a third region of the target object in the target coordinate system; mapping coordinates corresponding to the third region to the aims In the image frame, target coordinates are obtained; an area in the target image frame corresponding to the target coordinates is used as the second area.

Further, acquiring the third region of the target object in the target coordinate system comprises: acquiring a feature color value of the target object; using a difference between the color value of the target image frame and the feature color value as the first matrix Calculating the third region according to the first matrix.

Further, calculating the third region according to the first matrix includes: determining a first coordinate point according to a color value of an element of the first matrix, wherein a color value of the first coordinate point is the first a largest of the elements of the matrix; a rectangular frame centered on the first coordinate point, wherein an average of color values of elements of the edge of the rectangular frame and a color value of an element within the frame of the rectangular frame The average value has a preset value relationship; when the target object is in the rectangular frame, the area where the rectangular frame is located is taken as the third area.

Further, each element of the first matrix has three dimensions of red, green, and blue, and determining the first coordinate point according to the color value of the element of the first matrix includes: each element in the first matrix The color values of the three dimensions of red, green, and blue are added to obtain a second matrix, wherein each element of the second matrix has one dimension; a preset number of peripherals of each element in the second matrix The color values of the elements are averaged, the average value is used as the color value of the elements in the second matrix; and the color values of the elements in the second matrix whose color values are within the first preset range are assigned Zero, a third matrix is obtained; the coordinates of the element having the largest color value in the third matrix are taken as the first coordinate point.

Further, adding color values of three dimensions of red, green, and blue of each element in the first matrix to obtain a second matrix includes: red, green, and blue of the first element in the first matrix Multiplying the result of the color values of the three dimensions by the second element in the weight value change matrix to obtain the second matrix, wherein the first element is any one of the elements in the first matrix The coordinates of the second element in the weight value change matrix are the same as the coordinates of the first element in the first matrix, and the weight value change matrix is as follows:

Where R _c is the uniaxial direction decreasing ratio, (N1 _c , N2 _c ) is the coordinate of the center point of the image, x is the abscissa of the pixel of the image, and y is the ordinate of the pixel of the image.

According to another aspect of the embodiments of the present application, there is provided an image processing apparatus, including: a first acquiring unit, configured to acquire a color reference value of a target object in a target image frame, wherein the color reference value is the a color value of the color of the color that is the most frequently occurring in the preset area; the second acquiring unit is configured to obtain a pixel point formed by the difference between the target image frame and the color reference value that is less than or equal to a preset value. a first area; a processing unit, configured to use an edge of the first area as an outline of the target object.

Further, the first acquiring unit includes: a selecting subunit, configured to select a second area where the target object is located from the target image frame; and a first acquiring subunit, configured to acquire the second area The central point is a plurality of pixel points in the preset area of the origin; a search subunit is configured to search for color values corresponding to the most frequently occurring color from the plurality of pixel points; and determine a subunit for The color value corresponding to the most frequently occurring color is used as the color reference value.

Further, the processing unit includes: a second acquiring subunit, configured to acquire a matrix formed by color values of all pixel points in the first region, where each element in the matrix represents a color of one pixel point a value; a calculation subunit, configured to calculate a gradient of the matrix, and calculate an absolute value of the gradient; a processing subunit configured to correspond an absolute value of the gradient to the target image frame to obtain the target The outline of the object.

Further, the selecting subunit includes: an acquiring module, configured to acquire a third area of the target object in a target coordinate system; and a mapping module, configured to map coordinates corresponding to the third area to the target image In the frame, a target coordinate is obtained; and a determining module is configured to use an area in the target image frame corresponding to the target coordinate as the second area.

Further, the obtaining module includes: an obtaining submodule, configured to acquire a feature color value of the target object; and a determining submodule, configured to use a difference between a color value of the target image frame and the feature color value as a a matrix; a calculation submodule for calculating the third region according to the first matrix.

Further, the calculating submodule includes: a first determining large module, configured to determine a first coordinate point according to a color value of an element of the first matrix, wherein a color value of the first coordinate point is the first a largest of the elements of a matrix; processing a large module for making a rectangular frame centered on the first coordinate point, wherein an average of color values of elements of the edge of the rectangular frame and a frame of the rectangular frame a mean value relationship exists between the average values of the color values of the elements; determining a large module for determining whether the target object is in the rectangular frame; and second determining a large module for when the target object is in the When the rectangle is inside the frame, the area where the rectangular frame is located is taken as the third area.

Further, each element of the first matrix has three dimensions of red, green, and blue, and the first determining large module includes: a first computing small module, configured to use each element in the first matrix The color values of the three dimensions of red, green, and blue are added to obtain a second matrix, wherein each element of the second matrix has one dimension; and a second calculation small module is used for each of the second matrix The color values of the elements of the preset number of elements are averaged, and the average value is used as the color value of the elements in the second matrix; the small module is used to set the color value in the second matrix The color value of the element within a predetermined range is assigned a value of zero to obtain a third matrix; a small module is determined for using the coordinates of the element having the largest color value in the third matrix as the first coordinate point.

Further, the first calculation small module is specifically configured to: add a result of adding color values of three dimensions of red, green, and blue of the first element in the first matrix, and a number in the weight value change matrix Multiplying two elements to obtain the second matrix, wherein the first element is any one of the first matrices, and the coordinates of the second element in the weight value change matrix are The coordinates of an element in the first matrix are the same, and the weight value change matrix is as follows:

According to another aspect of an embodiment of the present application, there is also provided a robot comprising: the above image processing apparatus.

In the embodiment of the present application, by using the color value of the color having the most occurrence as the color reference value, the pixel point in the target image frame that is equal to or close to the color reference value is regarded as the pixel point of the target object, and the target is acquired. In the image frame, the region where the color value and the color reference value are equal or close to each other (the first region), the first region is processed, and the edge of the first region is used as the contour of the target object, and the accurate determination is realized. The technical effect of the outline of the target object in the video further solves the technical problem in the prior art that the contour of the target object in the video cannot be accurately determined.

DRAWINGS

The drawings described herein are intended to provide a further understanding of the present application, and are intended to be a part of this application. In the drawing:

1 is a flow chart of an image processing method according to an embodiment of the present invention;

2 is a schematic diagram of a second area in a target image frame in an embodiment of the present invention;

3 is a schematic view of a center point of a second region in an embodiment of the present invention;

4 is a schematic diagram of a preset area in an embodiment of the present invention;

FIG. 5 is a schematic diagram showing the result obtained by clearing the color values of a plurality of pixel points in the second region in the embodiment of the present invention; FIG.

Figure 6 is a schematic view of a palm outline in an embodiment of the present invention;

7 is a schematic diagram of an image frame before reducing accuracy and color magnitude in an embodiment of the present invention;

8 is a schematic diagram of an image frame after reducing accuracy and color magnitude in an embodiment of the present invention;

9 is a schematic diagram of a second image frame in an embodiment of the present invention;

10 is a schematic diagram of recursive color values of a first image frame in an embodiment of the present invention;

11 is a schematic diagram of an image frame in which a color value of a pixel is a first result in an embodiment of the present invention;

12 is a schematic diagram of an image frame in which a color value of a pixel is a second result in an embodiment of the present invention;

13 is a schematic diagram of an image frame corresponding to an absolute value of a gradient of a second color value in an embodiment of the present invention;

14 is a schematic diagram of an image frame corresponding to a gradient of a recursive color value of a first image frame in an embodiment of the present invention;

15 is a schematic diagram of an image frame in which a color value of a pixel is a third result in an embodiment of the present invention;

16 is a schematic diagram of an image frame corresponding to a pixel point storing a preset ratio in an embodiment of the present invention;

17 is a schematic diagram of an image frame in which a color value of a pixel is a fourth result in an embodiment of the present invention;

18 is a schematic diagram of an image frame in which a color value of a pixel point is a fifth result in an embodiment of the present invention;

19 is a schematic diagram of an image frame corresponding to a first matrix calculated by using a second weighting formula in a color value of a pixel in an embodiment of the present invention;

20 is a schematic diagram of an image frame represented by a second matrix obtained according to a first matrix in an embodiment of the present invention;

21 is a schematic diagram of an image frame obtained after using a weight value change matrix according to an embodiment of the present invention;

22 is a schematic diagram of an image frame after some pixel points are cleared in an embodiment of the present invention;

23 is a schematic diagram of an image frame obtained by changing a color value of an image frame in an embodiment of the present invention;

24 is a schematic diagram of an image frame obtained by replacing a color value of a pixel point with an average value of color values of peripheral pixel points in an embodiment of the present invention;

25 is a schematic diagram of an image frame in which a color value of a pixel point is a third matrix in an embodiment of the present invention;

26 is a schematic diagram of an image frame including a first coordinate point and a color value of a pixel point as a third matrix in an embodiment of the present invention;

Figure 27 is a schematic view showing the expansion of the rectangular frame to the first range in the embodiment of the present invention;

28 is a schematic diagram of shrinking a rectangular frame to a second range in an embodiment of the present invention;

Figure 29 is a schematic view of the embodiment of the present invention after the rectangular frame is removed;

30 is a schematic diagram of a plurality of rectangular frames in an embodiment of the present invention;

31 is a schematic diagram of an image frame in a video in which a human fist is a target object according to an embodiment of the present invention;

32 is a schematic diagram showing a result obtained by clearing a color value of a plurality of pixel points in a second region when a human fist is used as a target object;

Figure 33 is a schematic view showing the outline of a human fist in the embodiment of the present invention;

Figure 34 is a schematic diagram of an image processing apparatus according to an embodiment of the present invention.

detailed description

The technical solutions in the embodiments of the present application are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present application. It is an embodiment of the present application, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present application without departing from the inventive scope shall fall within the scope of the application.

It should be noted that the terms "first", "second" and the like in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or order. It is to be understood that the data so used may be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in a sequence other than those illustrated or described herein. In addition, the terms "comprises" and "comprises" and "the" and "the" are intended to cover a non-exclusive inclusion, for example, a process, method, system, product, or device that comprises a series of steps or units is not necessarily limited to Those steps or units may include other steps or units not explicitly listed or inherent to such processes, methods, products or devices.

First, the technical terms involved in the embodiments of the present application are explained as follows:

Color value: The value of a color in a color mode. For example, when using the RGB color mode, the white color value is (255, 255, 255), the red color value is (255, 0, 0), the green color value is (0, 255, 0), and the blue color value is (0). , 0, 255), the yellow color value is (255, 255, 0), and the black color value is (0, 0, 0).

Gradient: In the case of a univariate real-valued function, the gradient is the derivative, or, for a linear function, the gradient is the slope of the line. In vector calculus, the gradient of the scalar field is a vector field. The gradient at a point in the scalar field points to the fastest growing direction of the scalar field, and the length of the gradient is the maximum rate of change.

Matrix Gradient: Take a 3*3 matrix as an example to illustrate the algorithm of matrix gradient.

matrix

[x,y]=gradient (C), then

Where x is the lateral gradient of the matrix, and the value of the first column of x is the second column of matrix C minus the first column of matrix C, eg, -5=(2-7)/1, the second column of x The value is the second column of matrix C minus the value of the first column plus the third column minus the value of the second column divided by 2, for example -3=((2-7)+(1-2))/ 2. The value of the last column of X is equal to the last column of matrix C minus the value of the second to last column divided by 1, for example -1 = (1-2)/2.

In accordance with an embodiment of the present application, an embodiment of an image processing method is provided, it being noted that the steps illustrated in the flowchart of the figures may be performed in a computer system such as a set of computer executable instructions, and, although The logical order is shown in the flowcharts, but in some cases the steps shown or described may be performed in a different order than the ones described herein.

FIG. 1 is a flowchart of an image processing method according to an embodiment of the present invention. As shown in FIG. 1, the method includes the following steps:

Step S102: Acquire a color reference value of the target object in the target image frame, where the color reference value is a color value of a color in which the target object appears most frequently in the preset area.

Step S104: Acquire a first region formed by a pixel point in the target image frame whose difference from the color reference value is less than or equal to a preset value.

In step S106, the edge of the first area is taken as the outline of the target object.

The color reference value of the target object is a color value that is very close to the true color of the target object. It can also be said that the color value in the target image frame is equal to or closer to the color reference value (the difference is less than or equal to the preset value). The point is considered to be the pixel of the target object. The first area is a set of pixel points in the target image frame whose difference from the color reference value is less than or equal to a preset value. The edge of the first area is taken as the outline of the target object.

By using the color value of the most frequently occurring color as the color reference value, the pixel point in the target image frame whose color value is equal to or close to the color reference value is regarded as the pixel point of the target object, and the color value and color in the target image frame are obtained. The region where the reference value is equal or the set of relatively close pixel points (the first region) is processed, and the edge of the first region is used as the contour of the target object, which solves the problem that the target cannot be accurately determined in the prior art. The technical problem of the contour of the object achieves the technical effect of accurately determining the contour of the target object.

Optionally, acquiring the color reference value of the target object in the target image frame includes: selecting a second region where the target object is located from the target image frame; acquiring a plurality of pixels in the preset region with the center point of the second region as an origin Point; finds the color value corresponding to the most frequently occurring color from multiple pixels; uses the color value corresponding to the most frequently occurring color as the color reference value.

The preset area may be an area composed of a plurality of pixel points centered on the center point of the second area (the entire area shown in FIG. 2 is the second area) (the black dot in the palm shown in FIG. 3) (eg, In the area where the white box of the center of the palm is shown in FIG. 4, the color value corresponding to the most frequently occurring color is searched from the preset area, and the color value corresponding to the most frequently occurring color is used as the color reference value. For example, the preset area is an area consisting of 81 pixels centered on the center point of the second area, and the color value corresponding to the most frequently occurring color is searched from the 81 pixels, and the color number with the most occurrences will be displayed. The corresponding color value is used as the color reference value. The color value of the pixel in which the difference between the color value and the color reference value in the second region is not in the preset range is cleared, and the result after the clearing is as shown in FIG. 5.

Optionally, using the edge of the first area as the contour of the target object comprises: acquiring a matrix formed by color values of all pixel points in the first area, where each element in the matrix represents a color of one pixel point The value is calculated; the gradient of the matrix is calculated, and the absolute value of the gradient is calculated; the absolute value of the gradient is mapped to the target image frame to obtain the contour of the target object.

The matrix composed of the color values of all the pixels in the first region has a fast change in the color value at the edge position of the target object, and the rate of change is large, and the change in the color value at the non-edge position of the target object is slow, and the rate of change is small. Since the gradient reflects the rate of change of the function, by finding the gradient of the matrix formed by the color values of all the pixels in the first region, it can be found that the absolute value of the color value of some elements in the matrix obtained by the gradient is small, and another part of the element The absolute value of the color value is large, and the pixel represented by the element having a small absolute value of the color value is the pixel point of the non-edge position of the target object, and the pixel represented by the element having the larger absolute value of the color value is the target. The pixel point at the edge position of the object corresponds to the absolute value of the gradient in the target image frame, thus obtaining the outline of the target object.

The matrix formed by the color values of all the pixels in the first region is subjected to a gradient to obtain a matrix, and all the elements of the matrix are taken as absolute values to obtain an absolute value of the gradient, and the absolute value of the gradient is corresponding to the target image frame. , get the outline of the target object, as shown in Figure 6, the white line of the palm edge. Figure 6 is a schematic illustration of the outline of the palm of the embodiment of the present invention.

Optionally, selecting the second region where the target object is located from the target image frame includes: acquiring a third region of the target object in the target coordinate system; mapping the coordinates corresponding to the third region to the target image frame to obtain the target coordinate; The area in the target image frame corresponding to the target coordinates is taken as the second area.

Before performing the above step S102, the accuracy of the video file can be reduced first, so that the operation speed of the computer can be improved, and the accuracy of the video file can be reduced by selecting one point for every four points in the single axis, that is, the fourth point. , 8th, 12th, and 16th are selected, and other data are ignored. The color value of the video is then reduced (for example, from 256 colors to 64 colors) to facilitate comparison of color differences for different frames. When the method provided by the embodiment of the present invention is used, the third region of the target object in the low-precision image frame needs to be acquired, and the coordinates of the low-precision image frame are restored to the coordinates of the original image frame, which is used to indicate the third. The coordinates of the region are mapped to the target In the image frame, the target coordinates are obtained, and the region in the target image frame indicated by the target coordinates is taken as the second region, and the precision of the target image frame is the same as the accuracy of the original video file.

7 is a schematic diagram of an image frame before the accuracy and color magnitude are reduced in the embodiment of the present invention, and FIG. 8 is a schematic diagram of an image frame after reducing the precision and the color magnitude in the embodiment of the present invention.

Optionally, acquiring the third region of the target object in the target coordinate system comprises: acquiring a feature color value of the target object; using a difference between the color value of the target image frame and the feature color value as the first matrix; Three areas.

As an optional embodiment, the first matrix may be calculated by using the difference between the color value of the target image frame and the feature color value as the first matrix, that is, subtracting the feature color value from the matrix of the color values of the target image frame. The matrix obtained by the matrix is used as the first matrix. The feature color value can characterize the color of the target object, and the feature color value is significantly different from the color value of the background. Taking the palm of the target object as an example, the characteristic color value of the target object can be defined as the color of the palm: (250, 200, 125) ± (50, 50, 50).

As another optional embodiment, the first matrix may also be calculated using a first weighting formula, where the first weighting formula is: first matrix = first coefficient * first result + second coefficient * second result + first Third coefficient * third result + fourth coefficient * fourth result + fifth coefficient * fifth result, wherein at least one of the first coefficient, the second coefficient, the third coefficient, the fourth coefficient, and the fifth coefficient is not equal to zero.

The specific process of obtaining the first result, the second result, the third result, the fourth result, and the fifth result is described in detail below.

A video file is usually composed of a plurality of image frames, and the first image frame and the second image frame are any two adjacent image frames in the same video file. Each image frame is composed of several pixels, and the color value of each pixel constitutes the color value of the image frame. The color value of the pixel in the first image frame is the first color value, and the color value of the pixel in the second image frame is the second color value.

In general, the color values of the image frames can be represented by a matrix. For example, the color value of the image frame Ti is a matrix Mi, and the elements of the matrix Mi correspond to the color values of each pixel of the image frame Ti. Since the value of the color value is related to the color mode, in order to facilitate the description, in the embodiment of the present invention, the RGB color mode is adopted unless otherwise specified. It should be noted that the image processing method provided by the embodiment of the present invention can also be used in other color modes, and only needs to adjust the size of the color value according to the color mode. When using the RGB color mode, each element of the matrix M1 has three dimensions. For example, the element of the 10th row and the 20th column of the matrix M1 is (255, 255, 0), indicating that the image frame T1 is located at the 10th line and the 20th. The color value of the pixel of the column is (255, 255, 0).

The following relationship exists between the recursive color value of the N+1th image frame, the recursive color value of the N+1th image frame and the Nth image frame:

The recursive color value of the (N+1)th image frame = (N+1th image frame + R* recursive color value of the Nth image frame) / (1+R). R is the preset scale value, and R has a value range of (0, 1). For example, R can take 0.9. When the first image frame is the first frame image frame, the recursive color value of the first image frame is obtained according to the first color value.

Find the gradient on both the left and right sides of the equal sign of the equation to obtain:

The gradient of the recursive color value of the (N+1)th image frame = (gradient of the N+1th image frame + R* gradient of the recursive color value of the Nth image frame) / (1 + R).

When using a matrix to represent the color values of an image, the gradient for the matrix is detailed above.

The process of obtaining the first result is: making a difference between the second color value of the second image frame and the recursive color value of the first image frame, and obtaining an absolute value, to obtain a first result. 9 is a schematic diagram of a second image frame in an embodiment of the present invention. Figure 10 is a diagram showing the recursive color values of a first image frame in an embodiment of the present invention. 11 is a schematic diagram of an image frame in which a color value of a pixel is a first result in an embodiment of the present invention.

The process of obtaining the second result is: obtaining a difference between the second color value of the second image frame and the recursive color value of the first image frame, obtaining a gradient for the difference, and obtaining an absolute value to obtain a second result. FIG. 12 is a schematic diagram of an image frame in which the color value of a pixel is a second result in the embodiment of the present invention.

The process of obtaining the third result is: finding the absolute value of the gradient of the second color value (representing the image as shown in FIG. 13) and the gradient of the recursive color value of the first image frame (the gradient representation of the recursive color value is as shown in the figure The difference between the images shown in Fig. 14 and the absolute value (indicating the image shown in Fig. 15) is obtained, and a third result is obtained.

The process of obtaining the fourth result is: obtaining a gradient matrix of the difference between the second color value and the recursive color value of the first image frame, and saving the preset proportion of elements according to the order of the elements of the gradient matrix, the preservation pre-preservation The gradient of the new gradient matrix after the proportional element is taken to find the absolute value as the fourth result. That is, the difference between the second color value and the recursive color value of the first image frame is a matrix, and the gradient is obtained for the matrix, that is, the gradient matrix of the difference between the second color value and the recursive color value of the first image frame is obtained. The gradient matrix contains multiple elements, which arranges multiple elements from large to small, keeps the elements of the preset ratio unchanged, and the rest are cleared. For example, the preset ratio is the top 10%, and the remaining 90% of the elements are Cleared, after this processing, a new gradient matrix (representing the image shown in Figure 16) is obtained, and the gradient is obtained for the new gradient matrix to obtain another matrix (for convenience of description, using Mt), and the matrix Mt is All elements take an absolute value, resulting in a fourth result (representing the image shown in Figure 17).

The process of obtaining the fifth result is that the difference between the first color value and the feature color value of the first image frame is taken as a fifth result (representing the image as shown in FIG. 18).

When a matrix is used to represent the color value of an image frame, the recursive color value is also a matrix. Since the result of finding the gradient on the matrix is still a matrix, the result of the gradient of the gradient of the matrix is still a matrix, and the addition, subtraction, multiplication, and division of the matrix are performed. The result obtained is still a matrix, so the first result, the second result, the third result, the fourth result, and the fifth result are all matrices.

Since there is an error in acquiring any of the first result, the second result, the third result, the fourth result, and the fifth result, by calculating the first matrix using a plurality of weighting methods, the error can be reduced, and the calculation is improved. The precision value of a matrix.

In consideration of averaging the effects of the color value differences, the first matrix may be calculated using a second weighting formula, wherein the second weighting formula is: first matrix = first coefficient *log _a (first result) + second coefficient * Log _a (second result) + third coefficient * log _a (third result) + fourth coefficient * log _a (fourth result) + fifth coefficient * log _a (fifth result), where a can take 2 , e (the bottom of the natural logarithm), 5, 10 and so on. 19 is a schematic diagram of an image frame in which a color value of a pixel is a first matrix in an embodiment of the present invention.

Optionally, calculating the third region according to the first matrix includes: determining a first coordinate point according to a color value of an element of the first matrix, wherein a color value of the first coordinate point is the largest of the elements of the first matrix; A coordinate frame is centered as a rectangular frame, wherein the average value of the color values of the elements of the edge of the rectangular frame has a preset numerical relationship with the average value of the color values of the elements in the frame of the rectangular frame; determining whether the target object is in a rectangular frame Inside; if the target object is inside a rectangular box, use the area where the rectangle is located as the third area.

Optionally, each element of the first matrix has three dimensions of red, green, and blue, and determining the first coordinate point according to the color value of the element of the first matrix includes: red, green, and each element in the first matrix The color values of the three dimensions of blue are added to obtain a second matrix, wherein each element of the second matrix has one dimension; the color values of the preset number of elements of each element in the second matrix are averaged, Using the average value as the color value of the element in the second matrix; assigning the color value of the element in the second matrix whose color value is within the first preset range to zero, to obtain a third matrix; and maximizing the color value in the third matrix The coordinates of the element are taken as the first coordinate point.

As described above, the color values of the image frames may be represented by a matrix. For example, the color value of the image frame Ti is a matrix Mi, and the elements of the matrix Mi correspond to the color values of each pixel of the image frame Ti. When the RGB color mode is adopted, each element of the matrix Mi representing the color value of the image frame Ti has three dimensions of red, green, and blue, and each element of the first matrix also has three dimensions of red, green, and blue. In the embodiment of the present application, the color values of the three dimensions of red, green, and blue of each element of the first matrix are added to obtain a second matrix (as shown in FIG. 20). The following describes in detail how to obtain the second matrix from the first matrix.

For example, when the first matrix

When the second matrix C2 is found. The first matrix C1 is a matrix of 3 rows and 4 columns, a total of 12 elements, each of which has three dimensions of red, green and blue. The elements in the second row and the third column of the first matrix C1 are (21, 32, 0), and the values of the elements in the second row and the third column of the first matrix C1 are in the three dimensions of red, green, and blue. Add, that is, add 21, 32, and 0 to obtain 53, that is, the value of the element in the second row and the third column in the second matrix C2 is 53. The calculation of the other elements in the first matrix is the same, and finally the second matrix is obtained.

The second matrix C2 is a matrix of 3 rows and 4 columns, a total of 12 elements, each element having only one dimension.

In order to reduce the weight of the peripheral position of the image frame and increase the importance of the intermediate position of the image frame, as an optional embodiment, the color values of the three dimensions of the first element in the first matrix may be red, green and blue. The result of the addition is multiplied by the second element in the weight value change matrix to obtain a second matrix, wherein the first element is any one of the elements in the first matrix, and the coordinates of the second element in the weight value change matrix The coordinates of the first element are the same as those of the first element, and the weight value change matrix is as follows:

Where R _c is the uniaxial direction decreasing ratio, (N1 _c , N2 _c ) is the coordinate of the center point of the image, x is the abscissa of the pixel of the image frame, and y is the ordinate of the pixel of the image frame. 21 is a schematic diagram of an image frame obtained after using a weight value change matrix in an embodiment of the present invention.

When the color values of the target objects are relatively close, the image frame can also be processed by using the following method, for example, the palm of the palm is a large difference between the color of the palm and the color of the black and white. It can be considered that the pixel points in the image frame whose color values are very close to the black and white color values are not the pixels of the image constituting the palm, and therefore, the color values of these pixels can be cleared. Since the black color value is (0, 0, 0) and the white color value is (255, 255, 255), it can be set that when the color value of the element in the second matrix is less than 150 or greater than 675, the element is considered not to constitute a palm. The element of the image, which clears the color value of this element. FIG. 22 is a schematic diagram of an image frame after some pixel points are cleared in the embodiment of the present invention.

Then, use the next formula to change the color value of the image frame:

The color value of the N+1th image frame (after the change) = the color value of the N+1th image frame (before the change) +r* The color value of the Nth image frame (already filtered), where r is the preset Proportional value.

For example, when the color value of the third image frame is obtained, the color value of the third image frame is compared with the second image. The sum of the filtered color values *r of the frame as the changed color value of the third image frame; after obtaining the color value of the fourth image frame, the color value of the fourth image frame and the third The sum of the filtered color values *r of the image frame as the changed color value of the fourth image frame; after obtaining the color value of the fifth image frame, the color value of the fifth image frame and the fourth The sum of the filtered color values *r of the image frame as the changed color value of the 5th image frame.

23 is a schematic diagram of an image frame obtained by changing a color value of an image frame in an embodiment of the present invention.

Averaging the color values of the preset number of elements of each element in the second matrix, and using the average value as the color value of the elements in the second matrix, for example, when the preset number is 9, the elements in the second matrix will be a is the average of the color values of the nine elements of the three rows and three columns at the center, and replaces the obtained average value with the color value of the element a. Further, for example, when the preset number is 25, the color values of 25 elements of 5 rows and 5 columns centered on the element a are averaged, and the obtained average value is substituted for the coordinates of the element a.

Hypothetical matrix

Replace each element in matrix C3 (except for the elements of the first row, the first column, the last row, and the last column) with the average of the color values of the elements of the 3 rows and 3 columns centered on the element, and then Clear the values of the elements of the first row, the first column, the last row, and the last column to get the matrix C4. The element 12 of the third row and the third column in the matrix C3 is taken as an example. The color values of the nine elements of the three rows and three columns centered on the element 12 are 4, 9, 15, 8, 12, 11, 20, 30, and 40, and the average of the nine color values is obtained to obtain 16.56. Therefore, the color of the elements of the third row and third column of the matrix C4 has a value of 16.56. The calculation of the other elements of the matrix C4 and so on, and finally the matrix C4:

Figure 24 is a diagram showing an image frame obtained by replacing the color value of a pixel point with the average value of the color values of the peripheral pixel points in the embodiment of the present invention.

After each element in the second matrix is replaced, the color values of the replaced elements of the replaced second matrix are sorted from large to small, and the elements whose color values are within the first preset range are The color value is assigned a value of zero, and the color value of the element whose color value is not within the first preset range is subtracted from a certain preset value to obtain a third moment. Array. The third matrix is the color value obtained after the image frame is filtered.

There may be multiple first preset ranges. For example, the first preset range may be that the color values are located 90%, 80%, etc. in the ranking. The first preset range is taken as an example in which the color value is located in the back 80%. Assuming that the replaced second matrix is a matrix C4, all elements of the matrix C4 except the elements of the first row, the first column, the last row, and the last column are sorted according to the color values from large to small, and the color values are The 80% of the elements in the sort are assigned zero, and the two elements of the 20% of the elements in the sort are subtracted from the third element in the sort by 14.11 (the fourth line in matrix C4) The element in column 3), the third element has the smallest difference from the smallest of the two elements. A third matrix C5 is obtained. The image represented by the third matrix is as shown in FIG.

The coordinates of the element having the largest color value in the third matrix are taken as the first coordinate point, and the coordinates of the element 2.45 in the third row and the third column in the third matrix are taken as the first coordinate point, and the displayed image is as shown in FIG. Black dots.

In general, the number of pixels of an image frame is on the order of thousands or tens of thousands. If the set of color values of the pixels of the image frame is represented by a matrix, the matrix contains thousands or even tens of thousands of elements. The number of rows or columns of this matrix may exceed 100. However, in the embodiment of the present invention, the use of such a high-order matrix example is very inconvenient, so the method provided by the embodiment of the present invention can only be described by taking a low-order matrix as an example. It should be noted that the method provided by the embodiments of the present invention is completely applicable to the case of a high-order matrix.

The first coordinate point is taken as the center for the rectangular frame, and the rectangular frame is expanded from small to large (as shown in FIG. 27), and the expansion of the rectangular frame can be performed in various ways, for example, the first one: the first coordinate point As a center for a rectangular frame, the rectangular frame is expanded from small to large until the average value of the color values of the elements of the edge of the rectangular frame and the average value of the color values of the elements in the frame of the rectangular frame satisfy a preset numerical relationship (for example, The average value of the color values of the elements of the edge of the rectangular frame is equal to the product of the average value of the color values of the elements in the frame of the rectangular frame and a certain preset ratio value. At this time, the expansion of the rectangular frame is stopped. Method 2: The first coordinate point is taken as the center as a rectangular frame, and the rectangular frame is expanded from small to large. During the expansion process, as long as an element whose color value is greater than 0 is found, the rectangular frame is expanded until the rectangular frame contains the element. After the expansion is completed, the rectangular frame is contracted (the contracted image is as shown in FIG. 28) until the average value of the color values of the elements of the edge of the rectangular frame and the average value of the color values of the elements in the frame of the rectangular frame satisfy. Preset numerical relationship.

After the expansion (or contraction) of the rectangular frame is completed, the information such as the center point, length, and height of the rectangular frame is saved. Then, the rectangular frame is cleared, and FIG. 29 is a schematic view after the rectangular frame is removed in the embodiment of the present invention. Use the same method to continue looking for the next rectangle. Figure 30 is a schematic illustration of a plurality of rectangular frames in an embodiment of the present invention. If a plurality of rectangular frames are found in total, the average value of the color values of the elements in the frame of each rectangular frame is determined, and the rectangular frame Km having the largest average value of the color values of the elements in the frame is filtered, and then the target object is judged whether In the rectangular box Km. It should be noted that, according to any two adjacent image frames (for example, the image frame T1 and the image frame T2), a rectangular frame Km can be obtained according to the image frame T1 and the image frame T2 according to the method provided by the embodiment of the present invention. . The rectangular frame Km determined by different adjacent image frames is also different. That is to say, a plurality of different rectangular frames Km can be obtained by processing different adjacent image frames.

It can be determined whether the target object is in the rectangular frame Km by determining whether the rectangular frame Km satisfies the preset condition. When the target object is in the rectangular frame Km, the rectangular frame Km is considered to be a valid rectangular frame, and the position and size information thereof are recorded. The preset condition may be any one or more of the following: (1) the average value of the color values of the elements in the frame of the rectangular frame is greater than a certain preset value; (2) the area of the rectangular frame is in a certain pre- Within the range; (3) the ratio of the length to the height of the rectangular frame is within a certain preset range; (4) the length or height of the rectangular frame is within a certain preset range; (5) the rectangle The total value of the color values of the elements in the box is greater than the average of the total value of the color values of the elements in the existing 10 valid rectangular boxes (for example, 85%); (6) The most frequently occurring occurrences in the rectangular frame The difference between the color value corresponding to the color and the average value of the color value corresponding to the most frequently occurring color in the existing 10 valid rectangular frames is within a certain range.

If it is determined that the target object is in the rectangular frame Km, the area in which the rectangular frame Km is located is taken as the third area.

When the target object is a human fist, since the color of the fist is relatively average, the method provided by the embodiment of the present invention can be used to determine the contour of the fist by using an image processing method similar to the target object as a palm. Specifically, the color reference value of the target object in the target image frame (as shown in FIG. 31) is acquired, where the color reference value is the color value of the color having the most occurrences in the preset area of the target object, for example, the preset area is An area consisting of 81 pixels centered on the center point of the second region, and the color value corresponding to the most frequently occurring color is searched from the 81 pixels, and the color value corresponding to the color having the most occurrence is used as the color. Reference. The color value of the pixel in which the difference between the color value and the color reference value in the second region is not in the preset range is cleared, and the result after the clearing is as shown in FIG. The first region of the pixel in the target image frame whose difference from the color reference value is less than or equal to the preset value is obtained, and the edge of the first region is used as the contour of the target object (as shown in FIG. 33).

According to an embodiment of the present invention, an image processing apparatus is also provided. The image processing apparatus can execute the image processing method described above, and the image processing method can be implemented by the image processing apparatus.

Figure 34 is a schematic diagram of an image processing apparatus according to an embodiment of the present invention. As shown in FIG. 34, the image processing apparatus includes a first acquisition unit 10, a second acquisition unit 20, and a processing unit 30.

The first obtaining unit 10 is configured to acquire a color reference value of the target object in the target image frame, where the color reference value is a color value of a color in which the target object appears most frequently in the preset area.

The second obtaining unit 20 is configured to acquire a first region formed by a pixel point in the target image frame that has a difference from the color reference value that is less than or equal to a preset value.

The processing unit 30 is configured to use an edge of the first area as a contour of the target object.

The first obtaining unit 10 uses the color value of the color having the most occurrence as the color reference value, and the pixel point in the target image frame that is equal to or close to the color reference value is regarded as the pixel point of the target object, and the second acquiring unit 20 Obtaining an area (first area) in which the set of pixel points whose color value is equal to or close to the color reference value in the target image frame, and the processing unit 30 processes the area where the set is located, and the edge of the first area is used as the target object. The contour solves the technical problem that the contour of the target object cannot be accurately determined in the prior art, and achieves the technical effect of accurately determining the contour of the target object.

Optionally, the first obtaining unit 10 includes: a selecting subunit, configured to select a second area where the target object is located from the target image frame; and a first acquiring subunit, configured to acquire the center point of the second area as an origin a plurality of pixels in the preset area; a search subunit for finding a color value corresponding to the most frequently occurring color from the plurality of pixels; determining a subunit for using the color corresponding to the most frequently occurring color The value is used as a color reference.

Optionally, the processing unit 30 includes: a second acquiring subunit, configured to acquire a matrix formed by color values of all pixel points in the first region, where each element in the matrix represents a color value of one pixel point; a unit for calculating a gradient of the matrix and calculating an absolute value of the gradient; a processing subunit for matching the absolute value of the gradient in the target image frame to obtain an outline of the target object.

Optionally, the selecting subunit includes: an obtaining module, configured to acquire a third area of the target object in the target coordinate system; and a mapping module, configured to map the coordinate corresponding to the third area into the target image frame to obtain the target coordinate; And a determining module, configured to use an area in the target image frame corresponding to the target coordinate as the second area.

Optionally, the obtaining module includes: an obtaining submodule, configured to acquire a feature color value of the target object; and a determining submodule, configured to use a difference between a color value of the target image frame and a color value of the feature as the first matrix; For calculating a third region according to the first matrix.

Optionally, the calculating submodule includes: a first determining large module, configured to determine a first coordinate point according to a color value of an element of the first matrix, wherein a color value of the first coordinate point is the largest of the elements of the first matrix Processing a large module for making a rectangular frame centered on the first coordinate point, wherein the average value of the color values of the elements of the edge of the rectangular frame and the average value of the color values of the elements in the frame of the rectangular frame have a preset value The relationship is determined by determining a large module for determining whether the target object is in a rectangular frame. The second determining large module is configured to use the area where the rectangular frame is located as the third area when the target object is in the rectangular frame.

Optionally, each element of the first matrix has three dimensions of red, green, and blue, and the first determining large module includes: a first computing small module, configured to red, green, and blue for each element in the first matrix The color values of the three dimensions are added to obtain a second matrix, wherein each element of the second matrix has one dimension; and a second calculation small module is used to preset a predetermined number of elements around each element in the second matrix The color values are averaged, and the average value is used as the color value of the elements in the second matrix; the small module is assigned to assign the color value of the element in the second matrix whose color value is within the first preset range to zero. Obtaining a third matrix; determining a small module for using the coordinates of the element having the largest color value in the third matrix as the first coordinate point.

Optionally, the first calculation small module is specifically configured to: add a result of adding color values of three dimensions of red, green, and blue of the first element in the first matrix, and the second element in the weight value change matrix Multiply, to obtain a second matrix, wherein the first element is any one of the elements in the first matrix, the coordinates of the second element in the weight value change matrix are the same as the coordinates of the first element in the first matrix, and the weight value change matrix is as follows :

According to an embodiment of the invention, a robot is also provided. The robot includes the image processing apparatus described above. The first acquiring unit 10 of the image processing apparatus uses the color value of the color having the most occurrence as the color reference value, and the pixel value in the target image frame that is equal to or close to the color reference value is regarded as the pixel point of the target object, The acquiring unit 20 acquires an area (first area) in which the set of pixel points whose color values and color reference values are equal or close to each other in the target image frame, and the processing unit 30 processes the area where the set is located, and the first area is processed. The edge as the contour of the target object solves the technical problem that the contour of the target object cannot be accurately determined in the prior art, and achieves the technical effect of accurately determining the contour of the target object.

In the above-mentioned embodiments of the present application, the descriptions of the various embodiments are different, and the parts that are not detailed in a certain embodiment can be referred to the related descriptions of other embodiments.

In the several embodiments provided by the present application, it should be understood that the disclosed technical contents may be implemented in other manners. The device embodiments described above are only schematic. For example, the division of the unit may be a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be combined or may be Integrate into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, unit or module, and may be electrical or otherwise.

The unit described as a separate component may or may not be physically separated as a unit display The components shown may or may not be physical units, ie may be located in one place or may be distributed over multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a standalone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application, in essence or the contribution to the prior art, or all or part of the technical solution may be embodied in the form of a software product stored in a storage medium. A number of instructions are included to cause a computer device (which may be a personal computer, server or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes: a U disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic disk, or an optical disk, and the like. .

The above description is only a preferred embodiment of the present application, and it should be noted that those skilled in the art can also make several improvements and retouchings without departing from the principles of the present application. It should be considered as the scope of protection of this application.

Claims

An image processing method, comprising:

Obtaining a color reference value of the target object in the target image frame, wherein the color reference value is a color value of a color of the target object having the most occurrences in the preset area;

Obtaining a first region formed by a pixel point in the target image frame that has a difference from the color reference value that is less than or equal to a preset value;

The edge of the first area is taken as the outline of the target object.
The method according to claim 1, wherein the obtaining a color reference value of the target object in the target image frame comprises:

Selecting a second region in which the target object is located from the target image frame;

Acquiring a plurality of pixel points in the preset area with the center point of the second area as an origin;

Finding a color value corresponding to the color having the most occurrences from the plurality of pixels;

The color value corresponding to the most frequently occurring color is used as the color reference value.
The method according to claim 2, wherein selecting the second region in which the target object is located from the target image frame comprises:

Obtaining a third area of the target object in the target coordinate system;

Mapping coordinates corresponding to the third region to the target image frame to obtain target coordinates;

An area in the target image frame corresponding to the target coordinate is used as the second area.
The method according to claim 3, wherein acquiring the third region of the target object in the target coordinate system comprises:

Obtaining a feature color value of the target object;

Taking a difference between a color value of the target image frame and the feature color value as a first matrix;

Calculating the third region according to the first matrix.
The method according to claim 4, wherein calculating the third region according to the first matrix comprises:

Determining a first coordinate point according to a color value of an element of the first matrix, wherein a color value of the first coordinate point is the largest of the elements of the first matrix;

Forming a rectangular frame centering on the first coordinate point, wherein an average value of color values of elements of edges of the rectangular frame has a preset numerical relationship with an average value of color values of elements in the frame of the rectangular frame ;

When the target object is in the rectangular frame, the area where the rectangular frame is located is taken as the third area.
The method according to claim 5, wherein each element of the first matrix has three dimensions of red, green, and blue, and determining the first coordinate point according to the color value of the element of the first matrix includes:

Adding color values of three dimensions of red, green, and blue of each element in the first matrix to obtain a second matrix, wherein each element of the second matrix has one dimension;

And averaging color values of a predetermined number of elements of each of the elements in the second matrix, and using the average value as a color value of an element in the second matrix;

And assigning a color value of an element of the second matrix whose color value is within a first preset range to zero, to obtain a third matrix;

The coordinates of the element having the largest color value in the third matrix are taken as the first coordinate point.
The method according to claim 6, wherein the color values of the three dimensions of red, green and blue of each element in the first matrix are added to obtain a second matrix comprising:

And multiplying a result of adding the color values of the three dimensions of the red, green, and blue colors of the first element in the first matrix, and multiplying the second element in the weight value changing matrix to obtain the second matrix, where The first element is any one of the first matrices, and the coordinates of the second element in the weight value change matrix are the same as the coordinates of the first element in the first matrix, The weight value change matrix is as follows:

Where R c is the uniaxial direction decreasing ratio, (N1 c , N2 c ) is the coordinate of the center point of the image, x is the abscissa of the pixel of the image, and y is the ordinate of the pixel of the image.
The method according to claim 1, wherein the edge of the first area is used as an outline of the target object comprises:

Obtaining a matrix formed by color values of all pixel points in the first region, wherein the matrix Each element represents the color value of a pixel;

Calculating a gradient of the matrix and calculating an absolute value of the gradient;

An absolute value of the gradient is mapped in the target image frame to obtain an outline of the target object.
An image processing apparatus, comprising:

a first acquiring unit, configured to acquire a color reference value of the target object in the target image frame, where the color reference value is a color value of a color of the target object having the most occurrences in the preset area;

a second acquiring unit, configured to acquire a first region formed by a pixel point in the target image frame that is smaller than or equal to a preset value of the color reference value;

And a processing unit, configured to use an edge of the first area as an outline of the target object.
The device according to claim 9, wherein the first obtaining unit comprises:

Selecting a subunit, configured to select a second region in which the target object is located from the target image frame;

a first acquiring subunit, configured to acquire a plurality of pixel points in the preset area with the center point of the second area as an origin;

Locating a subunit for finding a color value corresponding to a color having the most occurrences from the plurality of pixels;

Determining a subunit for using a color value corresponding to the most frequently occurring color as the color reference value.
The apparatus of claim 10 wherein said selecting subunit comprises:

An acquiring module, configured to acquire a third area of the target object in a target coordinate system;

a mapping module, configured to map coordinates corresponding to the third area into the target image frame to obtain target coordinates;

And a determining module, configured to use an area in the target image frame corresponding to the target coordinate as the second area.
The device according to claim 11, wherein the obtaining module comprises:

Obtaining a submodule, configured to acquire a feature color value of the target object;

a determining submodule, configured to use a difference between a color value of the target image frame and the feature color value as a first matrix;

a calculation submodule for calculating the third region according to the first matrix.
The device according to claim 12, wherein the calculation sub-module comprises:

a first determining large module, configured to determine a first coordinate point according to a color value of an element of the first matrix, wherein a color value of the first coordinate point is the largest of the elements of the first matrix;

Processing a large module for making a rectangular frame centering on the first coordinate point, wherein an average of color values of elements of edges of the rectangular frame and an average of color values of elements in the frame of the rectangular frame The value has a preset numerical relationship;

Determining a large module, configured to determine whether the target object is in the rectangular frame;

And a second determining large module, configured to: when the target object is in the rectangular frame, use an area where the rectangular frame is located as the third area.
The device according to claim 13, wherein each element of the first matrix has three dimensions of red, green and blue, and the first determining large module comprises:

a first calculation small module, configured to add color values of three dimensions of red, green, and blue of each element in the first matrix to obtain a second matrix, wherein each element of the second matrix has One dimension

a second calculation small module, configured to average a color value of a predetermined number of elements of each of the elements in the second matrix, and use the average value as a color value of an element in the second matrix;

An assignment small module, configured to assign a color value of an element of the second matrix whose color value is within a first preset range to zero, to obtain a third matrix;

Determining a small module for using the coordinates of the element having the largest color value in the third matrix as the first coordinate point.
The device according to claim 14, wherein the first computing module is specifically configured to:

And multiplying a result of adding the color values of the three dimensions of the red, green, and blue colors of the first element in the first matrix, and multiplying the second element in the weight value changing matrix to obtain the second matrix, where The first element is any one of the first matrices, and the coordinates of the second element in the weight value change matrix are the same as the coordinates of the first element in the first matrix, The weight value change matrix is as follows:

Where R c is the uniaxial direction decreasing ratio, (N1 c , N2 c ) is the coordinate of the center point of the image, x is the abscissa of the pixel of the image, and y is the ordinate of the pixel of the image.
The device according to claim 9, wherein the processing unit comprises:

a second acquiring subunit, configured to acquire a matrix formed by color values of all pixel points in the first region, where each element in the matrix represents a color value of one pixel point;

Calculating a subunit for calculating a gradient of the matrix and calculating an absolute value of the gradient;

Processing subunits for mapping an absolute value of the gradient in the target image frame to obtain an outline of the target object.
A robot comprising the image processing apparatus according to any one of claims 9 to 16.