color detection system for agricultural products based on electronic vision

development of a suitable method for color sorting of agricultural products such as: tomato, apple and orange, based on surface reflection and image formation was the goal of this research. a suitable scheme of color differentiation using an optic head and electrical processing circuitry for separating red from green and other colors is presented. a photo-conductive transducer, electrical resistance of which varied with different colors, was selected as the sensor element of the optic head. tests indicated that sensor resistance for blue and green was greater than that for yellow and red and furthermore, it increased with increasing distance of the object from the sensor. the system can easily be adjusted for other colors using potentiometer resistance change. thus, the designed system can be used for separating red tomatoes from green ones, green oranges from orange ones, red apples from yellow and green ones, etc. tests were conducted to check the system performance and to study the effect of some parameters such as type of fruit and ambient light intensity and so product shape and size on color detection accuracy of the system. results indicated that main and interaction effects of these parameters on detection accuracy were not significant and the system is capable of color differentiation under varying working conditions.