Click on “Download PDF” for the PDF version or on the title for the HTML version.
If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.
Sorting System Development of Potato Blackheart Based on Light Transmission Imaging
Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org
Citation: 2017 ASABE Annual International Meeting 1700648.(doi:10.13031/aim.201700648)
Authors: Fang Tian, Yankun Peng, Wensong Wei, Wenxiu Wang
Keywords: Potato blackheart, Nondestructive, Light transmission, Image processing, Potato classification.
Abstract. Potato blackheart is a kind of internal physiological disease which couldn‘t be found by people from the surface. In order to identify the potato blackheart without destruction and more accurately, a sorting system which could collect the light transmission images of potato tubers and automatically detect and sort the potatoes with blackheart was developed. The light sources used in this research were light emitting diodes and the wavelength was 705 nm. The blackheart part of potato tuber could weaken the transmitted light more effectively than the normal part, forming a dark-gray spot in the transmission image that could be detected by the computer image processing algorithm. During the image preprocessing period, the potato image area was extracted from the background, while median filtering method was used to reduce noise in the image. After that, the blackheart part of potato or other low grayscale area was separated based on the grayscale histogram. The parameters, namely the average grayscale value of the whole potato image area, the low grayscale area and the high grayscale area, the standard deviation of the whole potato image area, the low grayscale area and the high grayscale area were calculated based on the processed image. 72 potatoes were selected to validate the property of the system, and they were divided into a calibration group and a validation group according to the ratio of 3:1. After image preprocessing, MLR discriminant method was used to build a model to identify the samples. Eventually, approximately 89% of all the samples were detected and diverted accurately. The result showed that the nondestructive sorting system developed in this study could be used to detect the potatoes with blackheart accurately.
(Download PDF) (Export to EndNotes)
|