Texture-Based Estimation Technique For An Object
Fruit Recognition techniques have been developed based on color, shape, size attributes. But many of the fruit
images may have similar or identical color, shape, size values. Therefore using only these few features of color, shape and
size would be not sufficient and effective enough to identify and distinguish the fruit images. Hence a new fruit recognition
system is been proposed which will combine shape, color, size, texture and intensity to increase accuracy of recognition.
Index Terms—Fruit Recognition, Texture, Wavelet Transform, Co-occurrence Features, k nearest neighbor, color, SVM.