Fruit detection positioning and orchard map construction method based on machine vision

Abstract

The invention discloses a fruit detection positioning and orchard map construction method based on machine vision, and the method comprises the steps: employing a mobile robot carrying an RGBD camera to achieve the detection, counting and picking positioning of a fruit target, and fusing the information of a visual odometer and the point cloud information data of a laser radar; sLAM mapping and positioning in an orchard environment are realized by adopting a nonlinear optimization method; and fusing the two threads based on a multi-thread programming method so as to realize real-time fruit detection positioning, global mapping and self-positioning of the mobile robot applied to an actual orchard environment, and realize a global path planning function of the mobile robot in an orchard based on the constructed orchard semantic map. By means of the method, the application fields of intelligent orchard cruise, intelligent orchard yield estimation, picking robot orchard environment perception and the like can be achieved, and the traditional orchard planting industry can be better helped to be transformed and upgraded to the automation and intelligence direction.

Description

technical field [0001] The invention relates to the technical fields of target detection, visual SLAM, smart agriculture and picking robots, in particular to a method for fruit detection and positioning and orchard map construction based on machine vision. Background technique [0002] my country's fruit planting industry occupies a huge proportion of agricultural production in the Lingnan area. But nowadays fruit growing using traditional fruit growing methods is already facing big problems. First of all, the fruit picking operation is a time-consuming, labor-intensive, but low-efficiency link in the entire fruit planting and production relationship. At the same time, fruit picking itself has strong seasonal requirements, so it needs to be invested in a concentrated time. Intensive labor, which brings extremely high planting costs to agricultural producers. Secondly, under the background of my country's aging population and the reduction of agricultural labor force, the t...

AI Technical Summary

Problems solved by technology

However, the system detects and models fruit tree trunks and uses them as positioning markers. It does not use individual fruit semantic information that is more effective for picking operations. Its application performance can be further improved in actual orchard picking scenarios.

Images