Infrared weak target detection method based on visual contrast mechanism

Abstract

The invention discloses an infrared weak target detection method based on the visual contrast mechanism. The infrared weak target detection method based on the visual contrast mechanism comprises the following steps: (1) extracting a 8-direction gradient saliency map of the original infrared image; (2) performing binarization processing on the 8-direction gradient saliency map to obtain a gradient saliency binarization map; (3) performing optimizing processing on the gradient saliency binarization map according to the size feature of the weak infrared target and eliminating isolated noise points and large size background gradient saliency regions to obtain an optimized gradient saliency binarization map; (4) using the visual contrast mechanism to calculate the local contrast of the saliency region of the optimized gradient saliency binarization map, eliminating the false targets to obtain the gradient saliency binarization map with the visual contrast mechanism processing; (5) positioning the target in the original infrared image according to the gradient saliency binarization map with the visual contrast mechanism processing to complete the detection of the weak infrared target. The infrared weak target detection method based on the visual contrast mechanism solves the problem of low detection precision of weak infrared target with low signal to noise ratio.

Description

technical field [0001] The invention belongs to the technical field of digital image processing and target detection, and relates to an infrared weak and small target detection method based on a visual contrast mechanism. Background technique [0002] Infrared detection technology uses the temperature difference between the target and the background to identify the target. It has the advantages of strong anti-interference ability and good concealment. It has been widely used in military reconnaissance, missile guidance, intelligent security and other fields. However, with the increase of imaging distance and the interference of background factors, the targets in the infrared image are mostly presented as spots, and the signal-to-noise ratio of the image is low, and the target has no obvious shape or texture features. brought great difficulty. [0003] At present, infrared target detection methods can be divided into single-frame detection and multi-frame detection. The met...

AI Technical Summary

Problems solved by technology

Existing single-frame detection methods mainly include morphological filtering, K-means clustering, median filtering, background prediction, local maximum, correlation filter, wavelet transform, gradient operator, etc. Most of these single-frame detection methods are single Filter calculation and binarization processing, when it is applied to the detection of weak infrared targets with low signal-to-noise ratio, the detection rate is below 90%, and the false alarm rate is between 30% and 80%.

At present, there is another emerging infrared target detection method, that is, the target detection method based on the visual mechanism. This method is a simulation of the target screening mechanism in the human visual system. Attention mechanism, but the single-frame image processing time of this algorithm is more than 2s, and its main adaptation object is more significant large-scale targets

Images