Graph Domain Multi-Scale Infrared Dim Small Target Detection Method Based on Local Gradient Trilateral

Abstract

The invention discloses a graph-domain multi-scale infrared weak and small target detection method based on local gradient trilaterals. The method is as follows: converting an infrared image containing weak and small targets into a local gradient trilateral represented by a node and its edge weight relationship Graph signal; Secondly, according to the multi-scale transformation of the graph Laplacian matrix, the local gradient trilateral graph signal is multi-scale decomposed to obtain the low-frequency sub-bands of the graph signal at different scales; thirdly, according to the edge weight of the graph signal The relationship locally weights the high-frequency sub-bands of each scale, and takes the median value as the new central node coefficient, and then performs multiplicative fusion on the high-frequency sub-bands after local weighting; finally, the high-frequency sub-bands after multiplicative fusion The sub-band is segmented by adaptive threshold, the spatial position of the target is determined, and the detection result is output.

Description

technical field [0001] The invention belongs to the field of infrared image processing, and in particular relates to a multi-scale infrared dim and small target detection method based on local gradient trilateral. Background technique [0002] Infrared weak and small target detection technology is an extremely important key technology in the field of infrared imaging detection, and is widely used in automatic target recognition. When the target is far away from the infrared imaging detection system, the target not only has no obvious shape, structure, texture and other characteristic information in the obtained infrared image, but also has a relatively small imaging area, accounting for only a few to a dozen in the image. pixels. In addition, the complexity and harshness of the imaging environment and the limitations of the performance of the detection device make infrared images usually present poor visual effects such as low contrast, low signal-to-noise ratio, and blurre...

AI Technical Summary

Problems solved by technology

In addition, the complexity and harshness of the imaging environment and the limitations of the performance of the detection device make infrared images usually present poor visual effects such as low contrast, low signal-to-noise ratio, and blurred edges; Stable and effective detection of weak and small targets in infrared images with limited and complex background interference has become the main technical difficulty in the field of infrared imaging detection

[0003] Aiming at the difficult problem of infrared weak and small target detection in complex backgrounds, domestic and foreign scientific research scholars have carried out a lot of research work. According to the different objects to be processed, single-frame weak and small target detection methods can be divided into indirect detection methods based on background prediction and target feature-based methods. The direct extraction method; the detection method based on background prediction usually uses a filter or filter operator with better performance to predict and estimate the background component, and indirectly separates the weak and small Objective; commonly used classic methods include maximum mean / median filtering, morphological Top-Hat filtering, two-dimensional minimum mean square error filtering TDLMS, etc. Background adaptability is poor

[0004] In recent years, the method of directly extracting weak and small targets by using the local characteristics of the target has been favored by many researchers. Inspired by the mechanism of the human visual system, Chen et al. proposed a weak and small target detection method based on the local contrast difference between the target and the background. Establish the corresponding local contrast map, amplify the local intensity difference between the target and the background, and then select the threshold to extract the weak target; although this kind of method has good detection effect and certain robustness, it has high brightness and strong edge components for processing complex background, there will still be a high false alarm rate

Images