论文部分内容阅读
依据非制冷红外成像探测系统噪声分析结果建立噪声数学模型,以二元假设与极大似然比检验为基础,提出了应用于非平稳高斯噪声下的目标恒虚警率(CFAR)探测算法。该算法的目的是在无先验知识、不稳定的噪声环境中寻找已知外形的目标。经过与固定门限方法以及高斯模型二元假设恒虚警率探测算法比较之后得出,文中的算法在探测低信噪比小目标时有较高的探测概率,并且对噪声具有自适应性能,仿真实验验证了该方法的正确性和可行性。
Based on the noise analysis results of uncooled infrared imaging detection system, a mathematical model of noise was established. Based on binary hypothesis and maximum likelihood ratio test, a target CFAR detection algorithm was proposed for non-stationary Gaussian noise. The purpose of this algorithm is to look for the target of a known shape in a non-priori, unstable, noisy environment. After comparing with the fixed threshold method and the bivariate constant false alarm rate detection algorithm of Gaussian model, the algorithm in this paper has higher probing probability in detecting small targets with low signal-to-noise ratio, and has adaptive performance to noise. The simulation The experiment verifies the correctness and feasibility of the method.