	US 11,720,119 B1
	Path planning method of mobile robots based on image processing
An Cui, Changchun (CN); Tianmengyu Liang, Changchun (CN); Liyuan Liu, Changchun (CN); Yaohui Ma, Changchun (CN); Yingping Xu, Changchun (CN); and Mengmeng Yang, Changchun (CN)
Assigned to JILIN UNIVERSITY, Changchun (CN)
Filed by Jilin University, Changchun (CN)
Filed on Sep. 15, 2022, as Appl. No. 17/945,883.
Claims priority of application No. 202210087144.X (CN), filed on Jan. 25, 2022.
Int. Cl. G05D 1/02 (2020.01); G06T 7/11 (2017.01); G06T 7/136 (2017.01); G06T 7/60 (2017.01); G06T 7/73 (2017.01)

CPC G05D 1/0274 (2013.01) [G05D 1/0214 (2013.01); G06T 7/11 (2017.01); G06T 7/136 (2017.01); G06T 7/60 (2013.01); G06T 7/73 (2017.01); G06T 2207/20044 (2013.01)]

5 Claims

1. A path planning method of mobile robots based on image processing, comprising:

preprocessing of a map image: calculating a safety distance s between a mobile robot and a surrounding obstacle during a movement of the mobile robot based on external geometric features of the mobile robot, forming a circular range on the map image with a boundary point of the obstacle as a center and the safety distance s as a expansion radius to set a safety range, and marking the safety range; performing skeleton feature extraction on the map image after the marking to obtain a reference path map;

obtaining of an initial path: building an adjacency matrix, and performing path solution on the reference path map based on the adjacency matrix to obtain the initial path;

optimizing the initial path:

performing neighborhood expansion on the initial path to obtain an expansion path;

performing segmentation on the expansion path to obtain n numbers of segmented paths, and optimizing the n numbers of segmented paths separately;

splicing the n numbers of segmented paths after the optimizing to obtain an optimized full path;

wherein the performing segmentation on the expansion path to obtain n numbers of segmented paths, and optimizing the n numbers of segmented paths separately, comprises:

determining a segment length cd of each of the n numbers of segmented paths;

searching each of path points on the initial path and obtaining coordinates of the path points in each the segmented path;

calculating minimum and maximum values of horizontal and vertical coordinates in coordinate points, respectively, for each the segmented path, and the minimum and maximum values being denoted as: [i_min:i_max, j_min:j_max];

calculating a size of an area to be extracted by i_min, i_max, j_min, j_max;

w=i_max−i_min+1;

l=j_max−j_min+1;

where w is a width of an extracted area, and l is a length of the extracted area;

the extracted area for each the segmented path being: [i_min:i_max, j_min:j_max];

building an adjacency matrix for each the segmented path;

optimizing each the segmented path; and

obtaining an optimization result for the n numbers of segmented paths;

wherein the path planning method further comprises: controlling a mobile robot to move along the optimized full path.