一种分析噪声背景下目标特征有效性的方法

A Method Used to Analyse the Effectiveness of Target Features in Noisy Background

本文根据目标特征和噪声分布，在一定假设条件下推出了最小错误率（Ｂａｙｅｓ错误率）的表达式，从而把特征分布、噪声大小直接与分类的最小错误率联系起来，给出了一种分析噪声背景下目标特征的有效性的方法。

According to the distributions of target features and noise, this paper presents the formula of the minimum error probability(Bayes error probability) in some conditions. This formula relates directly the minimum error probability to the distributions of features and noise and can be used to analyse the effectiveness of target features in noisy background. As an example,the effectiveness of IR. radiation features of target and decoy is analysed in the paper.