Second order Statistical Texture Features from a New CSLBPGLCM for Ultrasound Kidney Images Retrieval

Chelladurai CALLINS CHRISTIYANA, Vayana Perumal RAJAMANI

Abstract


This work proposes a new method called Center Symmetric Local Binary Pattern Grey Level Co-occurrence Matrix (CSLBPGLCM) for the purpose of extracting second order statistical texture features in ultrasound kidney images. These features are then feed into ultrasound kidney images retrieval system for the point of medical applications. This new GLCM matrix combines the benefit of CSLBP and conventional GLCM. The main intention of this CSLBPGLCM is to reduce the number of grey levels in an image by not simply accumulating the grey levels but incorporating another statistical texture feature in it. The proposed approach is cautiously evaluated in ultrasound kidney images retrieval system and has been compared with conventional GLCM. It is experimentally proved that the proposed method increases the retrieval efficiency, accuracy and reduces the time complexity of ultrasound kidney images retrieval system by means of second order statistical texture features.

 


Keywords


CBIR; Texture; Second order statistics; GLCM; CSLBPGLCM; ultrasound kidney images; Recall; Precision

Full Text:

PDF
By clicking ‘Accept’, you agree to the use of cookies, and to the Privacy Policy on this website. Read our Privacy and Cookie Policies to find out more.


Appl Med Inform is published since 1995.