Local Binary Pattern Method and Feature Shape Extraction for Detecting Butterfly Image

Abstract

ABSTRACT:Research in the field of information retrieval especially on image processing is proliferating. Various methods are developed to be able to detect images optimally and produce better accuracy. The process of image detection can use the dataset that exists around us. In this research, we use butterflies dataset, since the butterfly has unique colors, patterns, and diverse shapes. Therefore, we use local binary pattern method for texture feature extraction and region props for shape feature extraction. The results of each texture feature extraction and shape feature extraction will be a merging process. The results of the merging process get an accuracy of 66%. In addition, the system testing process with confusion matrix will produce 67.1% precision value, 66% recall and f-measure of 66.5%. The merging process of both methods shows the interplay of texture and shape extraction.

Fine-Grained Butterfly Classification in Ecological Images Using Squeeze-And-Excitation and Spatial Attention Modules

“…The k-nearest neighbor algorithm (KNN) is adopted as the butterfly classification algorithm. DSY Kartika et al [6] combine local binary pattern (LBP) and region attributes to extract texture feature and shape feature, respectively. Then, the texture feature extraction and shape feature are combined to classify the butterfly.…”