Research on pornographic images recognition method based on visual words in a compressed domain

摘要: In order to recognise and filter pornographic images, visual-word-based image representation has attracted more attention. An can be represented as a bag of visual words, which is analogous the bag-of-words text documents. However, most existing approaches create words from images in pixel domain, requires extra processing time decompress since are stored compressed formats. A novel recognition method based on domain proposed this study. There four steps method: (i) low-resolution constructed data; (ii) scale-invariant feature transform (SIFT) descriptors extracted image; (iii) vocabulary created SIFT descriptors; (iv) identified by using support vector machine (SVM) classifier. The experimental results indicate that accurately with much less computational time.