Cloth pattern recognition with four features (RSSM)

Automatic clothing pattern recognition is a challenging research problem due to rotation, scaling, illumination, and especially large intra-class pattern variations. This paper recognizes clothing patterns in four categories (plaid, striped, pattern-less, and irregular) and identifies clothing colors. To recognize clothing patterns, we propose a novel Radon Transform Descriptor (RTD), Scale Invariant Feature Transform (SIFT), Mathematical Morphology (MM) based global feature and a schema to extract Statistical Descriptor (STA) to capture global features of clothing patterns. They all are combined to recognize complex clothing patterns. Our approach achieves 96% recognition accuracy which significantly outperforms the state-of-the-art texture analysis methods on clothing pattern recognition.