Analysis of errors in fruit size inspecting based on machine vision

Research advancements on fruit size inspecting with machine vision was introduced. The relationship among fruit, camera lens and fruit image was studied based on geometric optic theory. Pixels in the fruit image stand for varied actual size because of varied positions of points on the fruit surface, which gave rise to calibration errors. The distance between camera and fruit was limited and not long enough to obtain the actual radius from the image, which caused radius errors. Since the location of the fruit relates to the core of the camera, the shape on the image was varied, which resulted in shape errors. Three formulas were set up to calculate the errors respectively.