Subway security check method based on combination of CPSNet and yolov3

Abstract

The invention discloses a subway security check method based on the combination of CPSNet (csresnextpanetspp) and yolov3, and the method comprises the steps: enabling the CPSNet to be integrated intothe yolov3 to form a new network structure, carrying out the training through employing a self-collected subway security check image, and marking and recognizing dangerous goods and auxiliary detection goods in a subway security check image. According to the invention, a novel neural network structure is used and intelligent identification of a complex subway security check image is realized.

Description

technical field [0001] The invention relates to a medical image bone segmentation method based on the combination of CPSNet and yolov3, and belongs to the field of computer image processing. Background technique [0002] The current target detection is divided into traditional method target detection and deep learning method target detection. [0003] Traditional object detection methods are mainly divided into three parts: sliding window, feature extraction and classifier classification. Given an image, use the sliding window method to traverse the entire image, extract candidate boxes, and then use algorithms in classic computer vision pattern recognition, such as color-based, shape-based, texture-based, etc. to each candidate box. In the classification process, if it is only a single target classification, it is only necessary to distinguish whether the object contained in the window is a target, and if it is a multi-target classification, it is necessary to Further dis...

AI Technical Summary

Problems solved by technology

Moreover, compared with other detection networks, yolov3 has greatly improved the detection speed, but the detection accuracy is relatively low. In order to ensure real-time and accuracy at the same time, a combination of attention mechanisms is used to form a new network for feature extraction and detection.

Images