An automated inspection station for machine-vision grading of potatoes

PC-based machine vision system for real-time computer-aided potato inspection

“…For instance, for a variety with rough surfaces, the SMR threshold value can be increased from 10% to 12% to offset the extra small mismatched areas between the ellipse and the original rough potato boundary. It is thus difficult to compare the accuracy of our algorithms with other published results because they did not provide enough information about their samples, such as variety and distribution, and our evaluation method is different from that used by others [1][2][3][4].…”

“…First, these algorithms are quite computation intensive and the throughput achievable is much lower than the necessary real-time speed. For example, a recently developed inspection system [4] could perform size and shape inspection at the speed of 3 potatoes/min, far below the real-time speed of 4 potatoes/s per lane for typical potato handling mechanical systems currently in use [5]. Second, the price of the commercially available machine vision systems is quite high, sometimes even exceeding the cost of the mechanical handling system.…”

“…Tao et al [1] developed algorithms to inspect potato shape and color, and Grenander and Manbeck [2] used stochastic models to describe potato shape and color. Neural networks were applied to classify potato color by Deck et al [3], and Heinemann et al [4] implemented their algorithm in a workstation to inspect potatoes for size and shape. We have identified two major factors preventing the routine use of these machine vision systems.…”

Minimalist AdaBoost for Blemish Identification in Potatoes

“…[5] used the HSI colour space to identify green potatoes as well as yellow and green apples. [6] graded potatoes by size and shape. In [7] a system uses green levels to detect green defects (greening and sprouting) in potatoes.…”

Grading of vegetative cuttings using computer vision

“…With an increasing need for automation, systems have been designed to grade produce such as fruit [1], vegetables [2] and shellfish [3]. Typically such systems are based on simple criteria such as weight or size that can be measured very quickly in a production line.…”