Colour texture classification for Human Tissue Images

This paper presents a promising automotive technique for the classification of colour human tumour and nontumour tissues. The proposed classification algorithm is based on texture feature colour fuzzy texture spectrum and two more features from colour Doppler spectra. SOMs classify the tissues, which got a classification accuracy of 99.92%. Compared with other works, our method achieves good classification results on colour texture images, demonstrating the performance of our proposal.