Borah et al. (2007) proposed a Wavelet Texture Analysis (WTA) based texture feature estimation technique for discriminating images of eight different grades of CTC (cutting, tearing, and curling) tea. This technique conjugates the feature information of one group of images along with the information of rest of the groups. This was executed by considering the range of different groups of images of the same granule size. The ranges were estimated using the existing statistical texture features, namely variance, entropy and energy, in difference form. Daubechies' wavelets transform (WT) based sub-band images were utilized for calculating these statistical features. For calculating the final feature set, a simplified version of Mahalanobis distance calculation was adopted. This provided the difference between two images in terms of texture variations. This dissimilarity measurement was carried out among the images of same group of the eight tea databases. For data visualization, principal component analysis (PCA) was used to visualize the existing classes of textures and distinguishable characteristics were observed among the new feature sets. Authors claimed that the unsupervised clustering algorithm viz. self organization map (SOM) successfully classified the images efficiently into appropriate clusters. Two neural networks, namely multi-layer perceptron (MLP) network and learning vector quantization (LVQ) used for texture classifications exhibited a classification accuracy of 74.67% and 80% in MLP and LVQ, respectively. The results received by MLP and LQV outperformed the results received by using existing statistical texture features. This algorithm mainly focused on analysis using tailor made uniform sized samples and did not consider other attributes such as presence of stalk and bold, etc., in the tea sample.6. Conclusion