Journal of Engineering and Applied Sciences
ISSN: Online 1818-7803ISSN: Print 1816-949x
Abstract
Leaf classification by using images based on their textures is the main objective of this study. Local Binary Pattern (LBP) operator is eminent extracting method but it is not effective especially in the cases where noise (noise occurs due to external sources and other reasons) in the images involved or corrupted the image patterns. Local Ternary Pattern (LTP) is another famous feature extracting method gives solution to some extent but not completely solves this problem. Towards achieving perfectness of classification by correcting noisy bits, we propose a method for both error detection and correction called Corrected LBP (CLBP) based on the analysis of uniform binary patterns which are appears more frequently in the natural images and almost all image structures. We suggested in our proposed method modification of bits in the pattern based on the analysis of neighbouring bits. It gives significant increase of accuracy and performance levels.
How to cite this article:
Anil Kumar Muthevi and Ravi Babu Uppu. An Efficient Leaf (Texture) Classification using Local Binary
Pattern with Noise Correction.
DOI: https://doi.org/10.36478/jeasci.2017.5478.5484
URL: https://www.makhillpublications.co/view-article/1816-949x/jeasci.2017.5478.5484