Infrared image target detection method and device, equipment and storage medium

Abstract

The invention discloses an infrared image target detection method and device, equipment and a storage medium, and the method comprises the steps: establishing an infrared image data set, carrying out the data enhancement preprocessing, and generating a training sample set; according to operators supported by an embedded platform and computing power provided by the embedded platform, an infrared image target detection network is constructed, and a backbone network of the infrared image target detection network is constructed by adopting a focus image rearrangement module, a plurality of Ghost-depthWise convolution modules with residual structures and an SPP convolution module; the multi-scale target detection layer is constructed by combining an FPN network with a PAN network; training the network by using the training sample set; performing channel pruning on the trained network and retraining; and converting the retrained network according to the requirements of the target platform. In this way, complex operator calculation is avoided, redundant feature information is abandoned, the representation content of the features is enriched, the module conversion problem does not exist, the operation speed is high, and the detection precision is high.

Description

technical field [0001] The invention relates to the field of infrared image processing, in particular to an infrared image target detection method, device, equipment and storage medium. Background technique [0002] In recent years, infrared imaging technology has been widely used in the fields of automatic driving, intelligent security, remote sensing, and industrial monitoring. In view of the demand for all-weather uninterrupted detection, more and more dual-light devices (visible light + infrared) are deployed. The demand for image processing technology is also gradually increasing. When the weather is good, the use of ordinary visible equipment can meet most of the needs. However, in rainy, foggy or nighttime conditions, the imaging effect of visible light equipment is poor, which is not conducive to the back-end image processing. The passive imaging principle of long-wave infrared is just right It can make up for the situation that visible light cannot obtain high-qual...

AI Technical Summary

Problems solved by technology

[0004] In view of this, the purpose of the present invention is to provide an infrared image target detection method, device, equipment and storage medium, which can solve the problem of complex human-vehicle target detection network model based on deep learning, too many parameters, slow calculation, and small infrared image detection. The problem with poor object detection

Images