A branch identification method based on standing tree occlusion of single image

Abstract

The invention provides a standing wood blocking branch identification method based on a single image, which relates to the forest investigation field. The method comprises the following steps: performing image pretreatment on a photographed standing wood image, then removing background information of the standing wood image through image segmentation; removing trunk parts to obtain a plurality of branch images; refining the image to obtain a plurality of branch skeletons; linearly fitting all the branch skeletons to obtain linear equations of all the branch skeletons; working out the similarity of arbitrary two branch skeletons which are corresponding to two branches according to the intercepts and the slopes of the linear equations of the arbitrary two branch skeletons; when the similarity is larger than a set value, making a judgment that the two branches are blocking branches. The method of the invention has the advantages of judging whether the branches are the blocking branches, and reducing the mis-operation of a pruning robot.

Description

technical field [0001] The invention relates to forest resources investigation, in particular to a single image-based recognition method for branches blocked by standing trees. Background technique [0002] Forest is an important renewable resource on the earth and a necessary condition for human survival. Forest provides a habitat for a variety of organisms and is the most important life support system on earth. And trees are the basic components of forests. In reality, in order to make trees grow better, they need to be pruned. [0003] At present, the pruning operation is generally done manually, but the operation efficiency is low, the labor intensity is high, the labor cost is high, and there are certain hidden dangers. In order to solve the above problems, the prior art provides a remote-controlled automatic tree pruning machine. The pruning machine can be manually operated and remotely controlled to perform pruning operations, but it is necessary to manually look up ...

AI Technical Summary

Problems solved by technology

[0005] The pruning robot in the prior art automatically recognizes the branches of standing trees through the vision system, and when the vision system of the pruning robot takes an image, it will also shoot the branches of other trees around the tree to be pruned together in the image. The branches of other trees are called occluded branches; and the current pruning robots cannot distinguish which are occluded branches by taking images, especially through a single image, which causes many misoperations of pruning robots

Images