Image retrieval system originally based on the text in the image retrieval process, but the system of the text -based image retrieval is not optimal because it takes a long time to give a description of each file and the human perception varies in interpreting or annotate an image so that the image retrieval based on queries the text is not in accordance with the desired. Therefore, the necessary image retrieval system that uses information characteristics (color, shape and texture) of the image itself. Ornamental fish is an animal that has a kind of very much other than it has a unique ornamental fish also varied, unique ornamental fish can be seen from the color, shape and texture are very diverse. Image retrieval system for ornamental fish characteristic information using HSV histograms for color feature extraction, canny method to obtain information form and order two static methods to obtain texture information. Test results using recall and precision for image retrieval based on color feature extraction generate an average value of 100 %, the image search based on shape feature extraction generate an average value of 90 %, the image search based on texture feature extraction generate an average value of 90 % and Search Extraction characteristics of the image based on a combination of color, shape and texture generate an average value of 90 %.