PATTERN-RECOGNITION MODELS FOR SPECTRAL REFLECTANCE EVALUATION OF APPLE BLEMISHES

Surface blemishes of various apple varieties were analyzed by their re flectance characteristics between 460 and 1130 nm. Normalized reflecta nce data were collected at 10 nm increments with liquid crystal tunabl e filters. Data were utilized as input values for various pattern reco gnition models specifically multi-layer back propagation, unimodal Gau ssian, K-nearest neighbor and nearest cluster algorithms. Partitioning data into 50:50 training and test sets, correct classification in sep arating unflawed versus blemished areas ranged from 62 to 96% (Year I) and from 73 to 85% (Year II). The algorithm which yielded the highest correct classification was the multi-layer back propagation but minor variation was found for number of hidden nodes or neural net architec ture. (C) 1998 Elsevier Science B.V. All rights reserved.