Image Based Search Engine-Like System for Monitoring Fauna Flora

Content Based Image Retrieval method will be used to match the image of the query with the existing image in the database and or the image on the internet. Similarity measurements are performed using the Euclidean distance function. The image to be used is an image with JPEG format. The use of image or image feature in searching for image in a database and or internet can not be avoided, this is because searching the image by using keyword or text is very biased and the result is far from expected. Search engine-like to monitor the decline in numbers and detect the presence of new types of fauna and flora in Indonesia. Image searching was Carrried out by using color and shape features. Search engine-like is expected also to be developed into image based search engine using CBIR method. In this work image searching was deployed using color and shape features. We used not less than 5, 000 flora (plants) and 5, 000 fauna (wildlife) images. The experimetns show that in term of precission and recall, effectivity of retrieval are 60 % and 30 % respectively.