Visual perception-based structure analysis of images for digital collection retrieval

We proposed an approach of visual perception-based image index and its retrieval by multiform query, such as query-by-sample, query-by-perception-words, or query-by-impression-words. By using eigen SGLD (space gray level dependence) matrices, the statistical and textural features of images for index were extracted, and the associations between these features and semantic features based on human visual perception were analyzed. Based on the analysis results, the scheme for imagery-based image retrieval by multiform query was proposed, and user satisfaction-based evaluation was made.