Efficient Stamps Classification by Means of Point Distance Histogram and Discrete Cosine Transform

摘要: The problem of stamp recognition addressed here involves a multi-stage approach which includes detection, localization and segmentation, features extraction finally, classification. In this paper we focus on the two last stages, namely by means Point Distance Histogram Discrete Cosine Transform, classification employing distance calculation Euclidean metrics. first stage leads to automatic stamps segmentation has been described in several previous papers it is based mainly color segmentation. feature extractor works binary images employs polar representation points gathered histogram form, later reduced Transform. At stage, compact descriptors are compared according reference objects (class’ centers), closest class taken as answer. some results selected experiments real documents having different types stamps. A comparison with classical two-dimensional DCT calculated over also provided prove high discriminative power developed approach.