Automatic Color-based Image Recognition Technique using LAB Features and a Robust Unsupervised Clustering Algorithm

An automatic color-based image recognition approach is presented in this article. A set of digital images will be clustered in several classes on the color similarity basis. The images are featured using LAB color space. Then, the obtained color-based feature vectors are clustered using a novel automatic unsupervised classification algorithm based on validation indexes. Some experiments, performed using the developed recognition technique, are also described in this paper. Key-Words: image recognition, color-based feature vectors, LAB color space, unsupervised clustering, region-growing, validation indexes.