Abstract
Texture features play an important role in most image retrieval techniques to obtain results of high accuracy. In this work, the face image retrieval method considering texture analysis and statistical features has been proposed. Textile features can also be extracted using the GLCM tool. In this research, the GLCM calculation method involves two phases, first: some of the previous image processing techniques work together to get the best results to determine the big object of the face image (center of face image) then, the gray level co-occurrence matrix GLCM is computed for gray face image and then some statistical texture features with second-order are extracted. In the second phase, the facial texture features are retrieved by finding the minimum distance between texture features of an unknown face image with the texture features of face images that are stored in the database system. The experimental results show that the proposed method is capable to achieve high accuracy degree in face image retrieval.
Highlights
The texture is a property that clarifies the structure of an image or defined as an organized recurrence of pixels or pattern on the exterior. it is complex optical patterns that are collections of pixels, or objects within patterns with the properties of bright effect, colors, shapes, etc. [1, 2] In order to understand the images of the human face, it is important to understand the characteristics of the face and how it appears
Texture analysis recognized the spatial variation of image pattern based on some mathematical operations and some mathematical models to obtained information from it and statistical models of assess of texture that considers the spatial relationship of pixels is the gray-level co-occurrence matrix (GLCM)[5]
Edge detection technique is used to represent the data in 0/1 that gained from the threshold operation on the gray image, a technique detects pixels at edges of image to obtained of big region, some of preprocessing operations like enhancement, thinning and limitation for B/W image are work together for extracting of texture features at end of first phase in our method, the texture features are extracted based on GLCM features for face object resulted from preprocessing operations
Summary
The texture is a property that clarifies the structure of an image or defined as an organized recurrence of pixels or pattern on the exterior. it is complex optical patterns that are collections of pixels, or objects within patterns with the properties of bright effect, colors, shapes, etc. [1, 2] In order to understand the images of the human face, it is important to understand the characteristics of the face and how it appears. When a face image is capture at a different time, and the skin texture may have some new scars, in this case, we can be considering texture features From this standpoint, the texture features for face recognition was adopted in this research. The texture features for face recognition was adopted in this research In this proposed method, edge detection technique is used to represent the data in 0/1 that gained from the threshold operation on the gray image, a technique detects pixels at edges of image to obtained of big region (center of face image), some of preprocessing operations like enhancement, thinning and limitation for B/W image are work together for extracting of texture features at end of first phase in our method, the texture features are extracted based on GLCM features for face object resulted from preprocessing operations. The results of the first phase of the proposed method will be used in the process of retrieving the face image based on the minimum distance between the texture features of an unknown image with face images stored in the database system
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
