Volume estimation of walnut using image processing method based on perspective laws and similar elements

Volume determination of agricultural products is done in many reasons such as survey of quality and grading. In recent year’s use of nondestructive methods and fast machine vision have been discovered expanse and popularity by computer and digital cameras. This study tries for development and evaluation of the new image processing algorithm to error correction of natural perspective of imaging and cutting of similar elements and sum of their volume for volume estimation of product that in present study walnut was used. With this method some prevalent method such as weight of displaced water as reference method, geometric mean diameter and field of view surface to better evaluation of this method. The means of error of the geometric mean diameter, field of view surface and similar element method were 9.6%, 10.2% and 10.7%, respectively. The results showed in machine vision because of natural perspective of imaging, small objects (walnut) than larger ones produce more error for volume measurement.