Design and Test with a Tomato Identification System based on Visual Technologies

Abstract: In this paper, a boundary control-based algorithm Binary Quantum Particle Swarm Optimization (BQPSO) that considers quantum particle swarm with Delta potential well was used to determine otsu threshold. In the optimization, particles moved in the Delta potential well with the best position POPSIZE as center. The best threshold was determined by updating individual extremum of a single particle pbest and global extremum of particles swarm gbest to their good-enough fitness values, in order for image segmentation. As for profiles, random circle method was used to extract radius of fruit circle. With binocular vision system, a Fourier-transform algorithm was adopted to extract offsets of left and right tomato images, and by marking their sorting baseline, they were matched according to sequential consistency principle.