DocumentCode :
3062039
Title :
Boosting minimalist classifiers for blemish detection in potatoes
Author :
Barnes, Michael ; Duckett, Tom ; Cielniak, Grzegorz
Author_Institution :
Lincoln Univ., Lincoln, UK
fYear :
2009
fDate :
23-25 Nov. 2009
Firstpage :
397
Lastpage :
402
Abstract :
This paper introduces novel methods for detecting blemishes in potatoes using machine vision. After segmentation of the potato from the background, a pixel-wise classifier is trained to detect blemishes using features extracted from the image. A very large set of candidate features, based on statistical information relating to the colour and texture of the region surrounding a given pixel, is first extracted. Then an adaptive boosting algorithm (AdaBoost) is used to automatically select the best features for discriminating between blemishes and non-blemishes. With this approach, different features can be selected for different potato varieties, while also handling the natural variation in fresh produce due to different seasons, lighting conditions, etc. The results show that the method is able to build ¿minimalist¿ classifiers that optimise detection performance at low computational cost. In experiments, minimalist blemish detectors were trained for both white and red potato varieties, achieving 89.6% and 89.5% accuracy respectively.
Keywords :
computer vision; feature extraction; image resolution; object detection; statistical analysis; AdaBoost; adaptive boosting algorithm; blemish detection; feature extraction; image colour; image texture; machine vision; minimalist classifiers; pixel-wise classifier; potato blemish detection; statistical information; Boosting; Computational efficiency; Computer vision; Data mining; Detectors; Feature extraction; Image segmentation; Machine vision; Optimization methods; Pixel;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Image and Vision Computing New Zealand, 2009. IVCNZ '09. 24th International Conference
Conference_Location :
Wellington
ISSN :
2151-2205
Print_ISBN :
978-1-4244-4697-1
Electronic_ISBN :
2151-2205
Type :
conf
DOI :
10.1109/IVCNZ.2009.5378372
Filename :
5378372
Link To Document :
بازگشت