Automated identification of tomatoes with diseases using artificial neural network trained with genetic algorithms

Computer vision, genetic algorithm and artificial neural network technology were synthetically applied to automatically identify the tomatoes with diseases. First, the images of tomato were captured through computer vision system, then to identify empty tomatoes applying the round value and detect abnormal tomatoes applying the variation of fruit diameter. Second, artificial neural network trained with genetic algorithms was employed to conduct experimental research. The experiments show that the methods mentioned can accurately identify tomatoes shape and meet the requirement of the classification. The accuracy rate of the identification of tomatoes with diseases is up to 100%.