Texture feature extraction and classification based on improved LBP
Meiju Liu, Feng Zhang · 2023
The traditional LBP (Local Binary Pattern) algorithm has the problems of poor sensitivity to noise and low robustness to manic point interference when extracting texture features. An Improved Local Noise Robustness Binary Pattern (ILNRBP) is proposed in this paper. Using this algorithm, the gray value of pixels in each field is replaced by the average value of four neighboring pixels, and the difference between the gray value and the mean value of the central pixel is calculated. Then, the binary value of the difference between the difference and the threshold value T is quantized to get the binary string, and the texture characteristic value of the corresponding defect is obtained. The obtained texture feature vectors are sent to the SVM model for classification, and the purpose of classification is finally achieved. The experimental results show that the improved method integrates the texture direction information into the texture feature extraction, and the discrimination effect is more obvious, and the anti-noise robustness of texture classification is improved.