RCNN-based cattle herd target detection method and equipment

Abstract

The invention provides an RCNN-based cattle herd target detection method and equipment. The method comprises the steps of obtaining a current cattle herd image; building a neural network target detection model, wherein the neural network target detection model comprises a feature extraction backbone network, a feature pyramid and a cascade detector, expansion deformable convolution is added in the feature extraction backbone network, feature information of each layer is deeply fused in combination with the feature pyramid, and an improved neural network target detection model is obtained; and inputting the obtained current cattle herd image into the neural network target detection model to detect the to-be-detected target object to obtain a detection result. According to the invention, the interference of shielding on cattle body characteristics is reduced, the contrast ratio with cattle house background information is increased, the cattle body characteristics of shielding cattle herds are enhanced, and the spatial resolution is improved, so that the cattle herd detection accuracy is improved, the cattle farm management efficiency is ensured, the cattle herd number is accurately checked, and the cattle herd number is monitored in real time through a visual means.

Description

technical field [0001] The present application relates to the field of computers, in particular to a cattle target detection method and equipment based on Expansion-Deformable Convolutional Networks-Cascade Recurrent Convolutional Neural Networks (E-DCN-Cascade RCNN). Background technique [0002] In the existing technology, in recent years, with the development of the animal husbandry industry in the direction of large-scale, informatization, and refinement, intensive cattle farms will gradually replace small-scale farming models such as retail farming. In large-scale cattle farms, in order to achieve refined management of individual cattle and to track the health status of each cattle, it is necessary to establish and improve the quality traceability system, and the key lies in the identification of individual cattle. The premise of identification is to detect the target of the cattle group, and then identify the individual cattle when the target detection has a high enoug...

AI Technical Summary

Problems solved by technology

[0005] One purpose of this application is to provide an RCNN-based cattle target detection method and equipment to solve the problem of how to improve the accuracy of cattle target detection in the prior art, visualize and monitor the situation of cattle in real time so as to realize efficient management of cattle farms question

Images