Obstacle recognition method and apparatus, and device, medium and weeding robot

A method for determining whether an obstacle exists in a candidate weeding region of a lawn so that a weeding robot may work in the lawn and avoid working at the obstacle, the method comprising the steps of:

• obtaining a candidate weeding region image;
• determining a candidate obstacle region in the candidate weeding region image according to color information of the candidate weeding region image;
• obtaining contour information of the candidate obstacle region, wherein the contour information includes chrominance information, range information, and roughness information;
• and
• determining, according to preset contour information determination conditions and the contour information, whether there is an obstacle in the candidate weeding region image, so that the weeding robot may work in the lawn and avoid working at the obstacle,
• wherein the step of determining, according to preset contour information determination conditions and the contour information, whether there is an obstacle in the candidate weeding region image comprises determining a value of the chrominance information, and then:
• a. if the value of the chrominance information is greater than or equal to a preset chrominance threshold, determining whether the range information is greater than a first preset range threshold, and whether the roughness information is less than a first preset roughness threshold;
• and
• if so, determining that there is an obstacle in the candidate weeding region image;
• and
• b. if the value of the chrominance information is less than the preset chrominance threshold, determining whether the range information is greater than a second preset range threshold, and whether the roughness information is less than a second preset roughness threshold;
• and
• if so, determining that there is an obstacle in the candidate weeding region image, wherein the first preset range threshold is less than the second preset range threshold.