CN109345551B - Method, system and computer storage medium for detecting concave envelope in outer contour of image - Google Patents

Abstract

A method, system and computer storage medium for detecting concave envelope in the outer contour of an image. The detection method includes: sampling the pixel points on the outer contour of the corroded image, connecting the sampling points two by two to form line segments, and obtaining an intersection with the outer contour of the input image after expansion to obtain the intersecting contour; Perform position correction, connect the corrected start and end positions of each contour to form a closed area, calculate the area of the closed area, and find the closed area with an area greater than a certain threshold, then the contour corresponding to the closed area is the concave envelope of the image ; Finally, the concave area of the concave envelope of the input image is closed and filled. The method has the characteristics of high precision and strong robustness, and can effectively realize the detection and filling of the concave envelope in the outer contour of the image.

Description

Method, system and computer storage medium for detecting concave envelope in outer contour of image

technical field

The invention relates to the technical field of image processing, in particular to a method, a system and a computer storage medium for detecting a concave envelope in an outer contour of an image.

Background technique

Spraying robot is an important coating production equipment, mainly used for metal and non-metal surface coating work, is the product of the organic combination of robotics and surface coating technology. Spraying robots are not only safe and environmentally friendly, but also can complete spraying tasks efficiently and with high quality.

The spraying trajectory of the spraying robot determines the spraying quality and efficiency of the workpiece. Before automatically planning the spraying trajectory of the target object, it is necessary to collect the image information of the target object through sensors such as cameras or light curtains. Due to the wide variety of target objects, there are many concave images in the collected images. The concave image is a special workpiece image, and the spraying trajectory planning needs to be carried out separately. Therefore, in image classification, all images need to be detected by concave envelope, which is the outer contour of the concave area in the image. If the image detection result is a concave image, the concave envelope is filled and then trajectory planning is done.

SUMMARY OF THE INVENTION

In order to solve the problem of spraying trajectory planning of concave images, the present invention provides a detection method, system and computer-readable storage medium for concave envelopes in outer contours of images. The detection method has the characteristics of high precision, strong robustness, etc. The detection of the concave envelope in the outer contour of the image is realized effectively.

A first aspect of the present invention provides a method for detecting a concave envelope in an outer contour of an image, comprising the following steps:

input image;

Extract the outer contour of the input image;

Corrode the input image to obtain the corroded image;

Extract the outer contour of the eroded image;

Superimpose the outer contour of the input image and the outer contour of the eroded image, that is, add the pixel values of the corresponding positions of the outer contours of the two images;

The pixel points on the outer contour of the eroded image are sampled at preset intervals, and the sampling points are connected in pairs to form a line segment, and then the intersection is obtained with the outer contour of the input image, and the intersected outer contour is recorded;

If there is no intersection, the input image does not include the outer contour of the concave envelope, and the input image is returned; if there is an intersection, the start position and end position of each intersecting outer contour are adjusted so that the outer contour and the start position sum The area of the closed area formed by the connecting line at the end position is the largest;

It is judged whether the area of the closed area is larger than a preset value, and if so, the outer contour corresponding to the area of the closed area is the concave envelope of the image.

In some specific embodiments, the step of inputting an image includes that the inputting image is a binarized two-dimensional image.

In some specific embodiments, before extracting the outer contour of the input image, a step of judging whether there is a hollow area in the input image is further included, and if there is, filling the hollow area.

In some specific embodiments, the pixel points on the outer contour of the eroded image are sampled at preset intervals, the sampling points are connected in pairs to form a line segment, and then an intersection is obtained with the outer contour of the input image, and the step of recording the intersected outer contour includes: :

Let the preset interval of sampling be L, set the number of all pixels on the outer contour of the corroded image to be N, when N/L has a remainder m, then the interval of the last segment is m, and the number of sampling points is n=fix (N/L)+1, where fix means rounding toward zero; when the remainder of N/L is 0, the number of sampling points n=N/L, where L, N, m, and n are all natural numbers;

Start _traversing from the sampling point H ₁ , connect the sampling point H ₁ with the subsequent _sampling points H ₂ ,... , connect the sampling point H ₂ and the subsequent _sampling points H ₃ ,... The outer contour of the image is intersected, and all the intersection points are recorded;

Determine whether the pixel points between adjacent intersection points are less than a predetermined value. If so, the outer contour where the adjacent intersection points are located belongs to the same outer contour; if not, the outer contour where the adjacent intersection points are located belongs to Outer contours of different segments;

Record all outer contours with intersection points.

In some specific embodiments, after the sampling points are connected to form a line segment, the line segment is expanded, and then an intersection is obtained with the outer contour of the input image.

In some specific embodiments, the step of adjusting the start position and the end position of the outer contour intersected by each segment so that the area of the enclosed area formed by the outer contour and the connecting line between the start position and the end position is the largest including:

Let the number of pixels on the outer contour intersected by each segment be r, and each pixel point is expressed as I ₁ , I ₂ ,..., I _r-1 , I _r , I ₁ is the starting position of the contour, and I _r is the end position of the contour;

Correct the start position I ₁ of the outer contour: keep the end position I _r of the outer contour unchanged, connect the start position and the end position of the outer contour, calculate the area of the closed area formed by the line connecting the outer contour and the start position and the end position, left and right Move the starting position until the area of the enclosed area is the largest, and update the starting position as I _b ;

Correct the end position I _r of the outer contour: keep the start position I _b of the outer contour unchanged, connect the start position and the end position of the outer contour, calculate the area of the closed area formed by the connection line between the outer contour and the start position and the end position, and move to the left and right End position until the area of the enclosed area is maximum, and update the end position to I _e .

In some specific embodiments, the method further includes correcting the start position _Ib of the outer contour again: keeping the end position Ie of the outer _contour unchanged, connecting the start position and the end position of the outer contour, and calculating the connection between the outer contour and the start position and the end position. The area of the closed area formed by the line, move the starting position left and right until the area of the closed area is the largest, and update the starting position to I _b' , then the outer contour between the starting position I _b' and the ending position I _e is the concave envelope of the image network.

In some specific embodiments, it also includes the step of filling the area where the concave envelope is located: drawing a rectangular frame according to the start position and the end position on the concave envelope, closing the area where the concave envelope is located, and Fill the enclosed area enclosed by the concave envelope and the rectangular box.

A second aspect of the present invention provides a detection system for a concave envelope in an outer contour of an image, the system comprising:

memory and one or more processors;

wherein the memory is connected in communication with the one or more processors, the memory stores instructions executable by the one or more processors, and the instructions are executed by the one or more processors , so that the one or more processors are used to perform the aforementioned method.

A third aspect of the present invention provides a computer-readable storage medium having computer-executable instructions stored thereon, which, when executed by a computing device, are operable to perform the aforementioned method.

The invention provides a method, a system and a computer storage medium for detecting a concave envelope in the outer contour of an image. The detection method includes: sampling the pixel points on the outer contour of the corroded image, connecting the sampling points two by two to form line segments, and obtaining an intersection with the outer contour of the input image after expansion to obtain the intersecting contour; Perform position correction, connect the corrected start and end positions of each contour to form a closed area, calculate the area of the closed area, and find the closed area with an area greater than a certain threshold, then the contour corresponding to the closed area is the concave envelope of the image ; Finally, the concave area of the concave envelope of the input image is closed and filled. The method has the characteristics of high precision and strong robustness, and can effectively realize the detection and filling of the concave envelope in the outer contour of the image.

Description of drawings

1 is a flowchart of a method for detecting a concave envelope in an image outer contour of the present invention;

2 is a schematic diagram of an input image;

3 is a schematic diagram of the outer contour of an input image;

FIG. 4 is a schematic diagram of the corrosion result of the input image;

5 is a schematic diagram of the outer contour of the corrosion image;

6 is a schematic diagram of image erosion;

7 is a schematic diagram of image expansion;

8 is a schematic diagram of the result of adding the outer contour of the input image and the outer contour of the eroded image;

9 is a schematic diagram of pixel sampling on the outer contour of the corroded image;

Figure 10 (a), (b), (c) is a schematic diagram of a line segment formed by connecting two sampling points;

11 is a schematic diagram of the intersection of the expanded line segment and the outer contour of the input image;

Figure 12 (a), (b), (c) are schematic diagrams before the contour start position correction;

Figure 13 (a), (b) are schematic diagrams of the correction result of the contour start position;

Figure 14 (a), (b) is a schematic diagram of the correction result of the contour end position;

Figure 15 (a), (b) is a schematic diagram of the contour start position correction result (correction again);

16 is a schematic diagram of a rectangular frame;

17 is a schematic diagram of a concave envelope of a closed image;

FIG. 18 is a schematic diagram of an image concave envelope filling image.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the specific embodiments and the accompanying drawings. It should be understood that these descriptions are exemplary only and are not intended to limit the scope of the invention. Also, in the following description, descriptions of well-known structures and techniques are omitted to avoid unnecessarily obscuring the concepts of the present invention.

In order to make the objectives, technical solutions and beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

A first aspect of the present invention provides a method 100 for detecting a concave envelope in an outer contour of an image, as shown in FIG. 1 , comprising the following steps:

Step 110: Input an image; specifically, the input image may be a binarized two-dimensional image InputImage, as shown in FIG. 2 . In addition to binarized images, the input images can also be other forms of digital images.

Step 120: Extract the outer contour of the input image.

Extract the outer contour Boundry of the input image, as shown in Figure 3. The outer contour of the input image is extracted by the sobel operator: the sobel operator contains two 3x3 convolution kernels Gx and Gy, where Gx=[-1,0,1;-2,0,2;-1,0, 1], Gy=[1,2,1;0,0,0;-1,-2,-1], each pixel in the image is convolved with these two convolution kernels, whichever is The output of the maximum value is the outer contour of the workpiece of the input image.

If the input image has a hollow area, you need to fill the hollow area first, and the pixel value of the hollow area can be set to 1 in the binary image.

Step 130: Corrosion processing is performed on the input image to obtain a corroded image.

Corrosion processing is performed on the input image InputImage to obtain the eroded image Img_erode, as shown in Figure 4.

Image erosion is a process of eliminating image boundary points and shrinking the image boundary inward, which can be used to eliminate small and meaningless objects. As shown in Figure 6, the corrosion of B to A can be expressed as AΘB. After image A is corroded by structural element B, image A is reduced by a circle, and its corrosion width is determined by the structural element. The structural element SE is determined by two parameters, shape and parameters. The shape parameter shape can be any shape such as a rectangle or a circle, and the parameters control the size and direction of the shape parameter shape. Usually, the structural element is smaller than the image to be processed. A two-dimensional structural element can be understood as a two-dimensional matrix, and the value of the matrix element is 0 or 1.

The specific steps of the image erosion are:

scan every pixel of the image with structuring elements;

The structuring element is "ANDed" with the binary image it covers (pixel value dot product at the corresponding position);

If all 1s, this pixel of the resulting image is 1, otherwise 0.

Let the structuring element be a 3*3 matrix SE=[1,1,1; 1,1,1; 1,1,1], a point P on the binary image, the surrounding pixels are P0, P1, P2, P3, P4, P5, P6, P7, form a matrix M=[P0, P1, P2; P3, P, P4; P5, P6, P7]. Matrix SE and matrix M are dot-multiplied to obtain matrix Q=[P0, P1, P2; P3, P, P4; P5, P6, P7]. When all elements in the matrix Q are 1, the point P is 1, otherwise it is 0.

Step 140: Extract the outer contour of the eroded image.

Extract the outer contour Boundry1 of the eroded image Img_erode, as shown in Figure 5. The outer contour of the eroded image Img_erode is extracted by the sobel operator.

Step 150: Superimpose the outer contour of the input image and the outer contour of the eroded image, that is, add the pixel values of the corresponding positions of the outer contours of the two images, as shown in FIG. 8 .

Step 160: Sampling the pixels on the outer contour of the eroded image at preset intervals, connecting the sampling points to form a line segment, and then finding an intersection with the outer contour of the input image, and recording the intersecting outer contour, as shown in Figures 9 and 10(a), (b) and (c).

Specifically, the pixel points on the outer contour of the corroded image are sampled, and the sampling interval is L. Let the number of all pixels on the outer contour of the eroded image be N, when N/L has a remainder m, the interval of the last segment is m, and the number of sampling points is n=fix(N/L)+1, where fix means Round to zero; when the remainder of N/L is 0, the number of sampling points n=N/L.

The sampling points are connected two by two to form a line segment and after expansion processing, the intersection is obtained with the outer contour Boundry of the input image, and the union of all the intersection points is obtained.

In a specific embodiment, it is assumed that the number of sampling points is n, and the serial numbers of the sampling points are 1, 2, 3, ..., n-1, n. Starting from the sampling point H ₁ , connect the sampling point H ₁ and the sampling point H ₂ to form a line segment H ₁ H ₂ and perform expansion processing, and then obtain the intersection of the expanded line segment H _1' H _2' and the outer contour Boundry of the input image , Obtain the intersection point matrix V=H _1' H _2' ∩ Boundry, the intersection point represents the point with the same coordinates in H _1' H _2' and Boundry, there may be more than one; then the sampling point H ₁ is connected with the sampling point H ₃ to form a line segment After H ₁ H ₃ and expansion processing, the expanded line segment H _1' H _3' and the outer contour Boundry of the input image are intersected, and the obtained intersection point matrix V ₁ =H _1' H _3' ∩ Boundry is merged into Matrix V; go down in sequence...; Finally, the sampling point H ₁ is connected with the sampling point H _n to form a line segment H ₁ H _n and after expansion processing, the intersection of the expanded line segment H _1' H _n' and the outer contour Boundry of the input image is obtained. , the obtained intersection point matrix V _n-2 =H _1' H _n' ∩Boundry is merged into the matrix V. When sampling point H ₁ traverses all sampling points (except itself), then starts traversing from sampling point H ₂ , and loops in turn, until sampling point H _n traverses all sampling points (except itself), and finally all intersection points are stored in the matrix V.

Determine whether the pixel points between adjacent intersection points are less than a predetermined value. If so, the outer contour where the adjacent intersection points are located belongs to the same outer contour; if not, the outer contour where the adjacent intersection points are located belongs to Outer contours of different segments; record all intersecting outer contours.

图像A经结构元素B膨胀后，图像A扩大了一圈，其膨胀宽度由结构元素决定。Dilation is the process of merging all background points in contact with an object into that object, expanding the boundary to the outside. As shown in Figure 7, the expansion of B to A can be expressed as

After the image A is expanded by the structuring element B, the image A is expanded by a circle, and its expansion width is determined by the structuring element.

In the step, the expansion processing of the image specifically includes:

scan every pixel of the image with structuring elements;

The structuring element is "ANDed" with the binary image it covers (pixel value dot product at the corresponding position);

If all 0s, this pixel of the resulting image is 0, otherwise 1.

Let the structuring element be a 3*3 matrix SE=[1,1,1; 1,1,1; 1,1,1], a point P on the binary image, the surrounding pixels are P0, P1, P2, P3, P4, P5, P6, P7, form a matrix M=[P0, P1, P2; P3, P, P4; P5, P6, P7]. Matrix SE and matrix M are dot-multiplied to obtain matrix Q=[P0, P1, P2; P3, P, P4; P5, P6, P7]. When all elements in the matrix Q are 0, the point P is 0, otherwise it is 1.

Step 170: Determine whether there are intersecting outer contours.

Find the coordinate position of each intersection point in the matrix V, and find its position number on the Boundry of the outer contour of the image, and arrange the position numbers in ascending order. When the difference between the position numbers is less than a certain threshold, it is classified as the same group.

If the matrix V is an empty set, indicating that there is no intersection point, the input image does not include the outer contour of the concave envelope, and the input image is returned, as shown in FIG. 3 .

If the matrix V is a non-empty set, it means that there is an intersection point, then adjust the start position and end position of the outer contour intersected by each segment, so that the area of the enclosed area formed by the outer contour and the connecting line between the start position and the end position is the largest. Because the concave envelope of the outer contour detected by the above steps may have a certain deviation from the actual concave envelope, as shown in Figure 12(b), the detected concave envelope is only a part of the actual concave envelope, so it needs to be fine-tuned Start and end positions Find the actual concave envelope start and end positions. The specific adjustment process is performed by placing the intersecting outer contour of each segment into the outer contour of the original input image, as shown in Figure 12-15. The specific steps are as follows:

Let the number of pixels on the outer contour intersected by each segment be r, and each pixel point is expressed as I ₁ , I ₂ ,..., I _r-1 , I _r ; I ₁ is the starting position of the outer contour, I _r is the end position of the outer contour;

Correct the start position I ₁ of the outer contour: keep the end position I _r of the outer contour unchanged, connect the start position and the end position of the outer contour, calculate the area of the closed area formed by the line connecting the outer contour and the start position and the end position, left and right Move the starting position until the area of the enclosed area is the largest, and update the starting position as I _b ;

Correct the end position I _r of the outer contour: keep the start position I _b of the outer contour unchanged, connect the start position and the end position of the outer contour, calculate the area of the closed area formed by the connection line between the outer contour and the start position and the end position, and move to the left and right End position until the area of the enclosed area is maximum, and update the end position to I _e .

In a specific embodiment, let the number of pixels on the outer contour intersected by each segment be r, and each pixel is sequentially represented as I ₁ , I ₂ , . . . , I _r-1 , I _r (I ₁ is The starting position of the contour, I _r is the ending position of the contour).

First, correct the start position I ₁ of the outer contour: keep the end position I _r of the contour unchanged, and move the start position left and right (for example, I _-1 and I ₃ separated by 2 position numbers, I _-1 is on the outer contour of the original input image pixel point), connect the start and end positions on the outer contour to obtain a closed area S _-1 composed of the chord I _-1 I _r and the outer contour I _-1 I _r , and the chord I ₁ I _r and the outer contour I ₁ The closed area S ₁ formed by I _r , the closed area S ₃ formed by the chord I ₃ I _r and the outer contour I ₃ I _r , compare the area sizes of the closed areas S _-1 , S ₁ , and S ₃ , as shown in Figure 12(a ) shown;

(a) If the area of the closed area S ₁ is the largest, reduce the sequence number interval to 1, the start position becomes I ₀ and I ₂ , connect the start and end positions of the outer contour, and obtain the relationship between the chord I ₀ I _r and the outer contour I The closed area S ₀ composed of ₀ I _r , the closed area S ₂ composed of the chord I ₂ I _r and the outer contour I ₂ I _r , compare the areas of the closed areas S ₀ , S ₁ , and S ₂ again, and find the one with the largest area. Close the area, and update the starting position as I _b ;

(b) If the area of the closed area S _-1 is the largest, then continue to move two serial number positions backward (or to the left), the start position becomes I _-3 , and connect the start and end positions of the outer contour to obtain the chord I _{- 3} I _r and the closed area S _-3 formed by the outer contour I _-3 I _r ;

(b1) When the area of the enclosed area S _-1 is larger than the area of the enclosed area S _-3 , reduce the interval of the position serial numbers to 1, and repeat the operation of (a);

(b2) When the area of the closed area S _-3 is larger than the area of the closed area S _-1 , continue to move the two serial number positions backward until the starting position with the largest area is found, then reduce the serial number interval to 1, and repeat (a) operation.

Then correct the end position I _r of the contour: keep the start position I _b of the outer contour unchanged, connect the start and end positions of the outer contour, move the end position left and right until the area of the closed area is the largest, and update the end position to I _e , refer to the relevant process (a)(b).

Finally, correct the start position _Ib of the contour again: keep the end position Ie of the outer _contour unchanged, connect the start and end positions of the outer contour, move the start position left and right until the area of the closed area is the largest, and update the start position to _Ib' , related to Refer to (a)(b) for the flow. Then the outer contour between the start position I _b' and the end position I _e is the concave envelope of the image.

Step 180: Determine whether the area of the closed area is greater than a preset value, if so, the outer contour corresponding to the area of the closed area is the concave envelope of the image; if not, the area corresponding to the closed area is The outer contour is not the concave envelope of the image.

Specifically, the modified start and end positions of each contour are connected to form a closed area, the area of the closed area is calculated, and the closed area with an area greater than a certain threshold is found. Then the contour corresponding to the closed area is the concave envelope Line of the image. As shown in Figure 11, there are 4 contour lines in total. Let the closed area composed of chord I _b' I _e and contour I _b' I _e be C ₀ , the closed area composed of chord I _b1' I _e1 and contour I _b1' I _e1 be C ₁ , and the closed area composed of chord I _b2' The closed area composed of I _e2 and contour I _b2' I _e2 is C ₂ , the closed area composed of chord I _b3' I _e3 and contour I _b3' I _e3 is C ₃ , find the closed area C whose area is greater than a certain threshold _i (i=0, 1, 2, 3), the concave envelope and concave area of the image can be obtained.

Further, the detection method further includes a step of filling the area where the concave envelope is located: drawing a rectangular frame according to the corrected start, end and end positions on the concave envelope to close the area where the concave envelope is located, and Fills the area where the concave envelope is located.

In some specific embodiments, a rectangular frame is drawn to enclose the concave area according to the modified start and end positions on the concave envelope, as shown in FIGS. 16 and 17 . Connect the start and end positions of the concave envelope, and draw a rectangular frame along the normal vector direction based on this to close the concave area, as shown in Figure 17; then fill the hollow area of the image (all the hollow areas are set to 1) so that the image becomes The fully enclosed image is shown in Figure 18; finally, the trajectory planning is carried out.

A second aspect of the present invention provides a detection system for a concave envelope in an outer contour of an image, the system comprising:

memory and one or more processors;

wherein the memory is connected in communication with the one or more processors, the memory stores instructions executable by the one or more processors, and the instructions are executed by the one or more processors , so that the one or more processors are used to perform the aforementioned method.

A third aspect of the present invention provides a computer-readable storage medium having computer-executable instructions stored thereon, which, when executed by a computing device, are operable to perform the aforementioned method.

To sum up, the present invention provides a method and system for detecting a concave envelope in an outer contour of an image, and a computer storage medium thereof. The detection method includes: sampling the pixel points on the outer contour of the corroded image, connecting the sampling points two by two to form line segments, and obtaining an intersection with the outer contour of the input image after expansion to obtain the intersecting contour; Perform position correction, connect the corrected start and end positions of each contour to form a closed area, calculate the area of the closed area, and find the closed area with an area greater than a certain threshold, then the contour corresponding to the closed area is the concave envelope of the image ; Finally, the concave area of the concave envelope of the input image is closed and filled. The method has the characteristics of high precision and strong robustness, and can effectively realize the detection and filling of the concave envelope in the outer contour of the image.

It should be understood that the above-mentioned specific embodiments of the present invention are only used to illustrate or explain the principle of the present invention, but not to limit the present invention. Therefore, any modifications, equivalent replacements, improvements, etc. made without departing from the spirit and scope of the present invention should be included within the protection scope of the present invention. Furthermore, the appended claims of this invention are intended to cover all changes and modifications that fall within the scope and boundaries of the appended claims, or the equivalents of such scope and boundaries.

Claims (10)

1. a detection method of concave envelope in the outer contour of an image, is characterized in that, comprises the following steps: input image; Extract the outer contour of the input image; Corrode the input image to obtain the corroded image; Extract the outer contour of the eroded image; Superimpose the outer contour of the input image and the outer contour of the eroded image, that is, add the pixel values of the corresponding positions of the outer contours of the two images; The pixel points on the outer contour of the eroded image are sampled at preset intervals, and the sampling points are connected in pairs to form a line segment, and then the intersection is obtained with the outer contour of the input image, and the intersected outer contour is recorded; If there is no intersection, the input image does not include the outer contour of the concave envelope, and the input image is returned; if there is an intersection, the start position and end position of each intersecting outer contour are adjusted so that the outer contour and the start position sum The area of the closed area formed by the connecting line at the end position is the largest; It is judged whether the area of the closed area is larger than a preset value, and if so, the outer contour corresponding to the area of the closed area is the concave envelope of the image. 2 . The method for detecting a concave envelope in an outer contour of an image according to claim 1 , wherein the step of inputting an image comprises that the input image is a binarized two-dimensional image. 3 . 3. the detection method of concave envelope in a kind of image outer contour according to claim 1, is characterized in that, before extracting the outer contour of input image, also comprises the step of judging whether there is hollow area in input image, if there is , the hollow area is filled. 4. The method for detecting a concave envelope in the outer contour of an image according to any one of claims 1-3, wherein the pixel points on the outer contour of the corroded image are sampled at preset intervals, and the sampling point After the two-by-two are connected to form line segments, an intersection is obtained with the outer contour of the input image, and the steps of recording the intersected outer contour include: Let the preset interval of sampling be L, set the number of all pixels on the outer contour of the corroded image to be N, when N/L has a remainder m, then the interval of the last segment is m, and the number of sampling points is n=fix (N/L)+1, where fix means rounding toward zero; when the remainder of N/L is 0, the number of sampling points n=N/L, where L, N, m, and n are all natural numbers; Start _traversing from the sampling point H ₁ , connect the sampling point H ₁ with the subsequent _sampling points H ₂ ,... , connect the sampling point H ₂ and the subsequent _sampling points H ₃ ,... The outer contour of the image is intersected, and all the intersection points are recorded; Determine whether the pixel points between adjacent intersection points are less than a predetermined value. If so, the outer contour where the adjacent intersection points are located belongs to the same outer contour; if not, the outer contour where the adjacent intersection points are located belongs to Outer contours of different segments; Record all outer contours with intersection points. 5 . The method for detecting a concave envelope in the outer contour of an image according to claim 4 , wherein after the sampling points are connected to form a line segment, the line segment is expanded and then combined with the input. 6 . Find the intersection of the outer contours of the image. 6. the detection method of the concave envelope in a kind of image outer contour according to claim 1, it is characterized in that, described adjusting the starting position and the ending position of the outer contour that each section intersects, makes the outer contour and the starting position and ending The steps with the largest area of the enclosed area formed by the location link include: Let the number of pixels on the outer contour intersected by each segment be r, and each pixel point is expressed as I ₁ , I ₂ ,..., I _r-1 , I _r , I ₁ is the starting position of the contour, and I _r is the end position of the contour; Correct the start position I ₁ of the outer contour: keep the end position I _r of the outer contour unchanged, connect the start position and the end position of the outer contour, calculate the area of the closed area formed by the line connecting the outer contour and the start position and the end position, left and right Move the starting position until the area of the enclosed area is the largest, and update the starting position as I _b ; Correct the end position I _r of the outer contour: keep the start position I _b of the outer contour unchanged, connect the start position and the end position of the outer contour, calculate the area of the closed area formed by the connection line between the outer contour and the start position and the end position, and move to the left and right End position until the area of the enclosed area is maximum, and update the end position to I _e . 7. the detection method of concave envelope in a kind of image outer contour according to claim 6, it is characterized in that, also comprise the starting position I _b of revising outer contour again: keep outer contour end position I _e constant, connect outer contour. The start position and end position of the contour, calculate the area of the enclosed area formed by the connection between the outer contour and the start position and end position, move the start position left and right until the area of the enclosed area is the largest, and update the start position to I _b ', then the start position The outer contour between I _b ' and the end position I _e is the concave envelope of the image. 8. The method for detecting a concave envelope in the outer contour of an image according to claim 7, further comprising a step of filling the area where the concave envelope is located: according to the start on the concave envelope Draw a rectangular frame for the position and end position, close the area where the concave envelope is located, and fill the enclosed area enclosed by the concave envelope and the rectangular frame. 9. A detection system for concave envelope in the outer contour of an image, characterized in that the system comprises: memory and one or more processors; wherein the memory is connected in communication with the one or more processors, the memory stores instructions executable by the one or more processors, and the instructions are executed by the one or more processors , so that the one or more processors are used to perform the method of any of claims 1-8. 10. A computer-readable storage medium having stored thereon computer-executable instructions that, when executed by a computing device, are operable to perform the method of any one of claims 1-8.

Images