An Integrated Color and Intensity Co-Occurrence Matrix for Batik Image Retrieval

Retrieving Batik Image with An Integrated Color and Intensity Co-Occurrence Matrix (ICCM) method is proposed in this research. This method extracts the texture and color feature of the Batik Image. Feature extraction using ICCM and similarity calculation with the Euclidian distance are required to retrieve the batik Image retrieval in the proposed model. There are three main steps in the ICCM for feature extraction, i.e., first, transform the RGB Batik image into HSV space, second, create the co-occurrence matrix from the hue and intensity value of HSV image, third, and update the co-occurrence matrix based on its weight. The Euclidian distance is used to measure the similarity between the query Batik image and Batik image in the database. The highest average precision is achieved when the quantization of HSV image is eight level of quantization. The highest average precision is 1, 1, 0.91, and 0.86 for the recall 0.25, 0.5, 0.75, and 1,. respectively.