Pedestrian detection method and device, equipment and medium

Abstract

The invention discloses a pedestrian detection method, device, equipment and medium, and the network structure of an improved YOLOv5 network, the improved YOLOv5 network comprises a backbone network, an improved back neck network, a feature fusion layer and a detection head, the improved back neck network is formed by stacking three FPN networks, the FPN network comprises a convolution layer, an up-sampling layer, a cross-layer splicing layer and a CSP module, and the CSP module comprises an upper sampling layer and a lower sampling layer. The other end of the cross-layer splicing layer is connected with the backbone network, and the input end of each feature fusion layer is respectively connected with a convolutional layer and a CSP module in a first FPN network, a CSP module in a second FPN network and a CSP module in a third FPN network in the improved neck network; training the improved YOLOv5 network through an infrared pedestrian training set to obtain a pedestrian detection model; pedestrian detection is performed through the pedestrian detection model to obtain a pedestrian detection result, and the technical problem of low detection precision of an existing pedestrian detection method is improved.

Description

technical field [0001] The present application relates to the technical field of image processing, and in particular to a pedestrian detection method, device, equipment and medium. Background technique [0002] Infrared thermal imaging technology is a passive non-contact optical imaging technology, which has the advantages of strong anti-interference ability, long working distance and day and night work, and is widely used in intelligent security systems. [0003] However, the contrast of infrared images is low, and the ability to distinguish details is poor, resulting in low detection accuracy of pedestrian detection methods in infrared surveillance videos, and false detections are prone to occur. Among them, the YOLOv5 network has gained widespread attention due to its good performance, but the large number of convolutional layers in its network structure makes it easy for low-resolution images to lose important feature information, resulting in low detection accuracy. C...

AI Technical Summary

Problems solved by technology

[0004] This application provides a pedestrian detection method, device, equipment and medium, which are used to improve the technical problem of low detection accuracy existing in existing pedestrian detection methods

Images