Stereoscopic vision navigation method based on green crop feature extraction invariance

Abstract

The invention belongs to the technical field of visual navigation, and discloses a stereoscopic visual navigation method based on green crop feature extraction invariance, which is used for extractinglocal features of green crops, and obtaining the change condition of the number of image feature points during angle change by using Gaussian homomorphic filtering and histogram equalization to process an image; establishing a crop elevation map by using the obtained feature points; extracting the edge contour of the target object in the image, and separating the green crop from the crop background; extracting navigation parameters from the confidence density map fused by the elevation map and the edge map, and adding navigation angle and side edge distance constraint conditions; and fittingthe navigation angle and the angle actually rotated by the agricultural robot, and verifying the correctness of a fitting function by adopting an S-shaped path. The agricultural robot platform is built, agricultural crop image information collected by the binocular camera is processed, parameters of the navigation process are extracted, and automatic navigation control over the agricultural robotis achieved.

Description

technical field [0001] The invention belongs to the technical field of visual navigation, and in particular relates to a stereo visual navigation method based on the invariance of green crop feature extraction. Background technique [0002] Currently, the closest existing technology: [0003] As a product of the development of modern agriculture, agricultural robots play a role in promoting the production and development of agriculture. Agricultural robots are composed of robots and automation technology, which get rid of traditional manual labor methods, save labor resources and improve agricultural production efficiency. According to recent statistics, about 795 million people worldwide are undernourished, an increase of 628 million compared with a decade ago. By 2050, the world population will reach about 9.1 billion, 34% more than today. At the same time, with the continuous acceleration of urbanization, the global urban population will increase from 49% to about 70%....

AI Technical Summary

Problems solved by technology

[0008] (2) When the natural light, the viewing angle of the image, etc. change, the acquisition of image information will be affected, resulting in low navigation accuracy of farmland robots

In order to better solve problems such as image matching, object recognition, and target tracking, many algorithms are improved for certain factors, and it is difficult to have a method that can solve the problem of mixing multiple factors.

Images