A method for optimizing robot dynamic targets based on three-threshold edge detection algorithm

Abstract

A method for optimizing a robot's dynamic target based on a three-threshold edge detection algorithm, comprising the following steps: firstly, obtaining a segmentation threshold T of a foreground and a background by using an Otsu threshold segmentation method, and introducing it into a Canny operator, defining a high threshold as T h , the lower threshold is T l； Iterate through the entire matrix, and use 8 connected regions to specifically divide the points determined as sub-weak edge pixels and weak edge pixels; when there are strong edge pixels in the 8 neighborhoods of the sub-weak edge pixels, the sub-weak edge pixels are changed into strong edge pixels and assigned a value of 1 (255); when there are one or more strong edge pixels or two or more sub-weak edge pixels in the 8 neighborhoods of the weak edge pixels, the weak edge pixels are changed into sub-weak edge pixels; finally, all non-strong edge pixels are set to 0. The present invention improves the determination effect of actual edges and false edges, and makes the detection results more accurate and complete.

Description

Technical Field

[0001] The invention relates to a method for optimizing a robot's dynamic target, and in particular to a method for optimizing a robot's dynamic target based on a three-threshold edge detection algorithm. Background Art

[0002] With the rapid development of the robotics industry, autonomous robots are increasingly being used. When dynamic obstacles appear in the working environment, the robot must first perform dynamic target detection before timely obstacle avoidance. The edge of an image is a basic feature and important information in an image. At the same time, image edge detection can effectively reduce the amount of data, eliminate a large amount of irrelevant information, and facilitate subsequent research. Summary of the invention

[0003] The purpose of the present invention is to provide a method for optimizing robot dynamic targets based on a three-threshold edge detection algorithm. The present invention adopts an improved Canny edge detection algorithm and improves the Canny operator to optimize the foreground object obtained by the traditional inter-frame difference method. On the basis of the traditional Canny edge detection algorithm, the double threshold detection is changed to a three-threshold detection, the judgment strategy is changed, and the judgment of the true edge, false edge and noise on the image is optimized. Through performance analysis, the scheme partially optimizes the effect of dynamic target detection, improves the accuracy of the existing edge detection algorithm, and makes the image edge more continuous.

[0004] The objective of the present invention is achieved through the following technical solutions:

[0005] A method for optimizing a robot's dynamic target based on a three-threshold edge detection algorithm, the method comprising the following steps:

[0006] First, the segmentation threshold T of the foreground and background is obtained by Otsu threshold segmentation method, and then introduced into the Canny operator, and the high threshold is defined as T h , the lower threshold is T l , where T h =0.6T,T l =1 / 3T h =0.2T;

[0007] On this basis, the middle threshold T is introduced m , so that T m =0.3T;

[0008] Iterate through the entire matrix and divide the points determined as secondary weak edge pixels and weak edge pixels into specific 8-connected regions; when there is a strong edge pixel in the 8-neighborhood of the secondary weak edge pixel, the secondary weak edge pixel is changed to a strong edge pixel and assigned a value of 1 (255); when there is one or more strong edge pixels or two or more secondary weak edge pixels in the 8-neighborhood of the weak edge pixel, the weak edge pixel is changed to a secondary weak edge pixel; finally, all non-strong edge pixels are set to 0.

[0009] The method for optimizing the dynamic target of a robot based on a three-threshold edge detection algorithm, wherein the three thresholds are a high threshold T h , a middle threshold T m , a low threshold T l .

[0010] The method for optimizing a robot's dynamic target based on a three-threshold edge detection algorithm, wherein the given three thresholds are:

[0011] 1) If the edge pixel gradient value is greater than the high threshold, it is determined to be a strong edge pixel, and the pixel value at this position is set to 1 (255);

[0012] 2) If the edge gradient value is less than the high threshold and greater than the middle threshold, it is determined to be a secondary weak edge pixel;

[0013] 3) If the edge gradient value is less than the middle threshold and greater than the low threshold, it is determined to be a weak edge pixel;

[0014] 4) If the gradient value of the edge pixel is less than the low threshold, it will be suppressed and the pixel at that position will be set to 0.

[0015] The advantages and effects of the present invention are:

[0016] The present invention takes into account the characteristics of edge detection, greatly reduces the amount of data, and eliminates information that can be considered irrelevant, retaining the important structural attributes of the image. The three-threshold detection improves the judgment effect of actual edges and false edges without significantly increasing the time complexity of the algorithm, making the detection result more accurate and complete. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a schematic diagram of gradient level classification of the present invention;

[0018] Figure 2 This is the edge screening flow chart of the three-threshold method of the present invention. DETAILED DESCRIPTION

[0019] The present invention will be described in detail below with reference to the embodiments shown in the accompanying drawings.

[0020] Among various edge detection algorithms, the Canny edge detection algorithm is the best. After non-maximum suppression, an image consisting of local gradient minima will be obtained. The image display is many discrete points, which basically represent the actual pixels in the image. However, there are still some edge pixels caused by noise or color changes. In order to solve the above problems, weak gradient values ​​are used to filter out these edge pixels that are not actual pixels, and edge pixels with high gradient values ​​are retained.

[0021] The present invention adds an intermediate threshold to the dual threshold detection step of the Canny edge detection algorithm. First, the segmentation threshold T of the foreground and background is obtained by the Otsu threshold segmentation method, and then introduced into the Canny operator, and the high threshold is defined as T h , the lower threshold is T l , where T h =0.6T,T l =1 / 3T h =0.2T.

[0022] On this basis, the middle threshold T is introduced m , so that T m =0.3T. The gradient distribution diagram is as follows Figure 1 shown.

[0023] For the given three thresholds, one is the high threshold T h , a middle threshold T m , a low threshold T l . 1. If the edge pixel gradient value is greater than the high threshold, it is determined to be a strong edge pixel, and the pixel value at this position is set to 1 (255). 2. If the edge gradient value is less than the high threshold and greater than the middle threshold, it is determined to be a sub-weak edge pixel. 3. If the edge gradient value is less than the middle threshold and greater than the low threshold, it is determined to be a weak edge pixel. 4. If the edge pixel gradient value is less than the low threshold, it will be suppressed, and the pixel at this position will be set to 0.

[0024] The three threshold detection steps include iteratively traversing the entire matrix, and using 8 connected regions to specifically divide the points determined to be sub-weak edge pixels and weak edge pixels. When there is a strong edge pixel in the 8 neighborhoods of the sub-weak edge pixel, the sub-weak edge pixel is changed to a strong edge pixel and assigned a value of 1 (255). When there is one or more strong edge pixels or two or more sub-weak edge pixels in the 8 neighborhoods of the weak edge pixel, the weak edge pixel is changed to a sub-weak edge pixel. Finally, all non-strong edge pixels are set to 0.

Claims

1. A method for optimizing robot dynamic targets based on a three-threshold edge detection algorithm, characterized in that: The method comprises the following steps: First, the segmentation threshold T of the foreground and background is obtained by Otsu threshold segmentation method, and then introduced into the Canny operator, and the high threshold is defined as T h , the lower threshold is T l , where T h =0.6T,T l =1 / 3T h =0.2T; On this basis, the middle threshold T is introduced m , so that T m =0.3T; Iterate through the entire matrix, and use 8-connected regions to divide the points determined as secondary weak edge pixels and weak edge pixels. When there is a strong edge pixel in the 8-neighborhood of the secondary weak edge pixel, the secondary weak edge pixel is changed to a strong edge pixel and assigned a value of 1. When there is one or more strong edge pixels or two or more secondary weak edge pixels in the 8-neighborhood of the weak edge pixel, the weak edge pixel is changed to a secondary weak edge pixel. Finally, all non-strong edge pixels are set to 0. 1) If the edge pixel gradient value is greater than the high threshold, it is determined to be a strong edge pixel, and the pixel value at this position is set to 1; 2) If the edge gradient value is less than the high threshold and greater than the middle threshold, it is determined to be a secondary weak edge pixel; 3) If the edge gradient value is less than the middle threshold and greater than the low threshold, it is determined to be a weak edge pixel; 4) If the gradient value of the edge pixel is less than the low threshold, it will be suppressed and the pixel at that position will be set to 0.

Images