Cabalan, Royneil Robert

Animal image recognition system using back propagation neural network - Cebu City CIT-U 2011

The project system is a recognition system that evaluates and recognizes a certain picture of a single animal in a side view position. The image of the animal is a colored one with a white background. The picture is loaded into the system and evaluated by preprocessing methods such as image cropping and image edge detection in order to come up with a boundary-lining image. The image will be fed on the neuronal network called back propagation for the recognition process (e.g. Chicken, Cat) and with corresponding animal sound. One application of the system is that, it could be used as an educational system for children. A child can also use it for him to check images and help him understand the different animals that can be found in the environment. The sample images of animals are usually of great differences in terms of their shape lining. The system uses the following types of animals: Bird, Cat, Chicken, Camel, Duck, Fish, Frog, Goat, Rabbit and Turtle.



001.6425 C319 2011