Weather Forecast Based on Color Cloud Image Recognition under the Combination of Local Image Descriptor and Histogram Selection

Numerous researchers have used machine vision in recent years to identify and categorize clouds according their volume, shape, thickness, height, coverage. Due the significant variations illumination, climate, distortion that frequently characterize cloud images as a type of naturally striated structure, Local Binary Patterns (LBP) descriptor its variants been proposed feature extraction methods for characterizing natural texture images. Rotation invariance, low processing complexity, resistance monotonous brightness are characteristics LBP. The disadvantage LBP is it produces binary data extremely noise-sensitive struggles on regions image “flat” because depends intensity differences. This paper considers Ternary (LTP) overcome drawbacks feature. We also propose fusion color characteristics, features, LTP features classification cloud/sky Morover, this study proposes apply Intra-Class Similarity (ICS) technique, histogram selection approach, with goal minimizing number histograms approach achieves better performance recognition less use by fusing using ICS technique choose potential histograms.