Matching Confidence Analysis Based on Neural Network for Edge Magnitude Cross Correlation
基于神经网络的边缘强度互相关匹配可信度分析

Matching confidence is an important measure to analyses the qu ality of image matching. For normalized edge magnitude cross correlation matchin g algorithm, the measure of matching confidence based on neural network is studi ed. The training samples used to train a BP network are the matching results wit h the reference image and several sensed images. The trained network can be adop ted to measure matching confidence. Experimetal results with real satellite imag es and aerial images prove the effectiveness of the method.