Three-dimensional Shape Recognition using a Charge-Simulation Method to Process Primary Image Features

Abstract The feasibility of the charge-simulation method (CSM) algorithm to process primary image features for three-dimensional shape recognition is examined. To achieve this, a machine vision system was developed which consists of a light source, light beam conditioner, artificial retina installed with photo-sensors, data transfer unit and a computer installed with analogue-to-digital converter peripherals. The system was used to acquire primary image features for oranges and eggplants. The features were transferred to a retina model identical to the prototype artificial retina and were compressed using the CSM by computing output signals at work cells located in the retina. With these signals, neural networks were trained to classify each image sample in order to identify their shape. An overall classification rate of 94·0% was obtained when the prototype artificial retina discriminated between distinct shapes of oranges and eggplants. An overall rate of 75% was achieved when discriminating between less distinct shapes of straight and curved eggplants. The results show that it is feasible for the artificial retina based on the CSM algorithm to process primary image features for three-dimensional shape recognition.