Weed recognition method and device, terminal equipment and readable storage medium

Abstract

The invention is suitable for the technical field of image processing, and provides a weed recognition method, which comprises the steps of obtaining a to-be-recognized image containing a plant group,segmenting the to-be-recognized image through a pre-trained target detection algorithm to obtain a pre-segmented image containing a key part of the plant group, and then clustering the pre-segmentedimage to obtain a clustered image; and obtaining more than one sub-image, obtaining the sub-image meeting a preset condition as a target image, and inputting the target image and the to-be-recognizedimage image into a pre-trained neural network model to obtain a weed type identification result. According to the invention, the to-be-recognized image image is segmented through the target detectionalgorithm, the segmented image containing the key parts of the plant group is obtained, weed classification and identification are carried out based on the segmented image containing the key parts, and whether human eyes can be simulated to identify the key parts of the plant group so as to realize the plant group classification effect is determined. The classification precision and the accuracy of the recognition effect are improved, weeds are prevented and controlled in the early stage, and economic losses caused by non-rice plants are reduced.

Description

technical field [0001] The present application belongs to the technical field of image processing, and in particular relates to a weed identification method, device, terminal equipment and readable storage medium. Background technique [0002] In the process of agricultural generation, how to distinguish and identify rice and weeds in the early stage is a very important link. [0003] However, due to the wide variety of weeds in rice fields, the similarity between different types is high and the gap is small, which is easy to confuse people. Therefore, the existing weed identification methods are difficult to accurately distinguish weeds, and thus cannot be realized in the early stage. The control of weeds is difficult to reduce the economic losses caused by weeds. Contents of the invention [0004] The embodiment of the present application provides a weed identification method, device, terminal equipment and readable storage medium, which can solve the problem that the e...

AI Technical Summary

Problems solved by technology

[0004] The embodiment of the present application provides a weed identification method, device, terminal equipment and readable storage medium, which can solve the problem that the existing weed identification method is difficult to accurately distinguish weeds

Images