The purpose of this research is to design a system for the classification of cape gooseberries based on their shape and color, using artificial vision, neural networks and image processing, as means to achieve this objective. It is a projective type of research, in turn its method is descriptive and its source is documentary. The image processing software and the neural network of the system is developed in the Matlab programming environment, the algorithm uses the principles of image processing to crop, segment, eliminate the background and filter the photo, to finally feed the ART neural network and proceed with the recognition of the characteristics of the fruit.