Extraction of Individual Perception Features for imagery-based Image Retrieval

In this paper, with the objective to retrieving digital collections more intuitively and flexible, we proposed an approach of compact feature extraction and retrieval of images which matched the human perceptual image similarity criteria. For this, the eigen and difference SGLD (space gray level dependence) matrices were used to extract the features of images. The associations of the extracted features with the individual´s perceptual similarity criteria regarding color and structure were analyzed. The user satisfaction-based evaluation illustrated the good performance of the proposed method