Geethalakshmi, S N2017-03-282017-03-282013https://ir.avinuty.ac.in/handle/avu/2256Fruit plays an important role in our life. India is exporting large number o f fruits to abroad in that mosambi is one among. Before exporting the fruits, the fruits have to be graded according to their quality. This paper presents an evaluation and comparison o f the performance o f three different extraction methods for classification of defect and non defect fruits. Three different feature extraction methods are GLCM (Grey Level Co-occurrence Matrix), shape features and intensity based features. The performance of each feature extraction method is evaluated and compared based on PNN (Probabilistic Neural Network) classifier. The experimental results suggest that shape feature outperformed than other two methods.