Dual-redundancy vision algorithm based on YOLOv5s and CSRT algorithm fusion

The invention discloses a dual-redundancy vision algorithm based on fusion of a YOLOv5s algorithm and a CSRT algorithm. The method comprises the following steps: firstly, taking an image shot by a loading camera as input, then carrying out nonlinear processing on the image by adopting a convolutional neural network, finally carrying out frame selection through a rectangular bounding box, marking a classification label on the recognized content, and carrying out recognition detection on an expected target; yOLOv5s is a model based on a deep neural network, and the position and category of an expected target are predicted by using a convolutional neural network; the CSRT algorithm performs target tracking through image feature extraction and superposition of a machine learning idea, and updates features online in real time according to changes of the target and the environment; the method specifically comprises the following steps: constructing a spatial constraint correlation filter; constructing a space relia