An Enhancement over Texture Feature Based Multiclass Image Classification under Unknown Noise
Abstract
In this paper we deal with classification of multiclass images using statistical texture
features with two approaches. One with statistical texture feature extraction of the whole image, another with feature extraction of image blocks. This paper presents an experimental assessment of classifier in terms of classification accuracy under different constraints of images. This paper examined classification accuracy of multiclass images without noise, with some unknown noise and after filtering of noise using feed forward neural network. Results shows that blocking of image improves the performance of classifier.
features with two approaches. One with statistical texture feature extraction of the whole image, another with feature extraction of image blocks. This paper presents an experimental assessment of classifier in terms of classification accuracy under different constraints of images. This paper examined classification accuracy of multiclass images without noise, with some unknown noise and after filtering of noise using feed forward neural network. Results shows that blocking of image improves the performance of classifier.
Keywords
statistical texture, feature extraction, pattern recognition, multiclass classification, neural network