Abstract
Texture-based analysis of images is a very common and much discussed issue in the fields of computer vision and image processing. Several methods have already been proposed to codify texture micro-patterns (texlets) in images. Most of these methods perform well when a given image is noise-free, but real world images contain different types of signal-independent as well as signal-dependent noises originated from different sources, even from the camera sensor itself. Hence, it is necessary to differentiate false textures appearing due to the noises, and thus, to achieve a reliable representation of texlets. In this proposal, we define an adaptive noise band (ANB) to approximate the amount of noise contamination around a pixel up to a certain extent. Based on this ANB, we generate reliable codes named noise tolerant ternary pattern (NTTP) to represent the texlets in an image. Extensive experiments on several datasets from renowned texture databases, such as the Outex and the Brodatz database, show that NTTP performs much better than the state-of-the-art methods.
Highlights
Noise is an inherent property of images and becomes a major obstacle in detecting texture patterns.Different types of noises might originate from different sources
Photons coming from an Sensors 2011, 11 object undergo a series of processing, and form a pixel in a CCD/CMOS sensor based digital camera, where there may be a possibility of inclusion of different types of noises in every stage, such as photon shot noise and readout noise
We use the well-known Outex texture database [31] to compare the performance of noise tolerant ternary pattern (NTTP) with four state-of-the-art methods, namely local binary pattern (LBP), LBP variance (LBPV), Completed local binary pattern (CLBP) and Local Ternary Pattern (LTP)
Summary
Noise is an inherent property of images and becomes a major obstacle in detecting texture patterns.Different types of noises might originate from different sources. Noise is an inherent property of images and becomes a major obstacle in detecting texture patterns. Over the last few decades several proposals have been proposed for detecting texture micro-patterns. A family of Gabor kernels (usually eight different orientations and five different scales) is convolved with an image to extract micro-patterns, such as lines and edges, in different scales and orientations. The statistics of such micro-patterns can be used to describe the underlying textures in the image, and has been used for different types of image based applications [4,5,6,7,8]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
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
