Image background clutter measurement method based on intrinsic derivation mechanism

Abstract

The invention discloses an image background clutter measurement method based on an intrinsic derivation mechanism. The method comprises the following steps: employing an autoregression model to simulate the intrinsic derivation mechanism of the brain, and decomposing a target image into an ordered part and a disordered part; for the ordered part, measuring the difference between the ordered part corresponding to the target image and a background clutter image by utilizing structural similarity; for the disordered part, analyzing the uncertainty of the structure of the disordered part by adopting a brightness adaptive local binary pattern, calculating the uncertainty of the structure by utilizing ILBP entropy, and then measuring the difference between the disordered part corresponding to the target image and the background clutter image in an ILBP entropy space; and combining the ordered part clutter measurement and the unordered part clutter measurement by adopting a standard varianceweighted combination strategy to obtain a final image background clutter measurement value. The target acquisition performance predicted by the method is highly consistent with the actual subjective target acquisition performance data, and is superior to the existing clutter measurement method in the aspects of correlation and root-mean-square error.

Description

technical field [0001] The present invention relates to the technical field of image processing, and relates to a method for measuring background clutter of visible light images, in particular to a method for measuring background clutter of images based on an intrinsic derivation mechanism, which is mainly used for evaluating the impact of target acquisition performance and camouflage patterns in photoelectric imaging systems The design and evaluation of clutter, quantitative evaluation of the impact of clutter on target detection and recognition, etc. Background technique [0002] With the improvement of the detector technology level and the progress of the production process, the sensitivity and resolution of the photoelectric imaging system have been greatly improved, and have reached or approached the background limit, which makes the background factor an important factor limiting the target acquisition performance of the photoelectric imaging system. factor. Therefore,...

AI Technical Summary

Problems solved by technology

Schmieder et al. proposed the statistical variance model SV, which is simple to calculate and easy to implement, but its representation and description of complex clutter images is not accurate enough; Biberman et al. proposed the edge probability clutter measure POE based on the sensitivity of the human eye to image edge features. The number of edge points in the image is used to measure the strength of clutter; however, this method is transitionally dependent on the image edge detection threshold, and there is no fixed standard for its selection, and the clutter quantification results obtained by different users are not comparable; Chang et al. The human eye is highly adaptive to image structure features, and measures clutter by calculating the similarity between the background and the target in terms of brightness, contrast and structure, and proposes a target structure similarity clutter measure TSSIM; this method uses The stability constant of the denominator is highly dependent, which makes the clutter calculation results unstable; Xu Dejiang et al. calculated the clutter based on the high adaptability of the human eye to the image structure and the image structure similarity measure SSIM, which is widely used in the image quality evaluation field. Structural similarity with the target, and using the principle of visual salience to weight the structural similarity, proposed the image structure difference clutter measure VSD; Xiao Chuanmin et al. considered that the human eye is sensitive to the image structure, and used the direction gradient histogram to represent the target structure feature, the Bhattachary coefficient is used to measure the difference between the background clutter and the target in the direction gradient histogram space, and the background clutter measurement method ESSIM that introduces the gradient distribution feature is proposed; this method pays more attention to the gradient structure characteristics of the target, and weakens the target The brightness information of

[0005] To sum up, the background clutter measurement methods based on human visual characteristics in the prior art do not make full use of the human visual perception model, and there are inaccurate descriptions of clutter representations, unstable calculation results, and dependence on specific types of images. problems such as large variability and the independence of each method

Images