A proposal of The Rareness Measure of pixel blocks and its application to region extraction

In this paper, we propose an effective measure to compute rareness' of pixel blocks compared with other blocks and also introduce a region segmentation method to extract significant blocks such as defects in visual inspection system. Our basic idea is based on the simple and realistic assumption that the target blocks should be detected must have different statistical characteristics from other blocks. In order to calculate the measure, we utilize three types of histograms of features; image intensity, gradient value and its direction. Some blocks which have extraordinary characteristics can automatically be detected by estimating relative distances between all blocks. Experimental results by actual images have proved that our method work well for extracting rare defect region without a priori knowledge about the defects.