Optical Remote Sensing Images Feature Extraction of Forest Regions

摘要

The forest region saliency extraction technology based on optical remote sensing image plays an important role in forest fire risk monitoring and forest area protection in the process of urbanization. In this paper, the large-area forest area in the optical remote sensing image is highlighted in the feature map, and the salient map is further obtained through the generated feature map to achieve accurate extraction of the optical remote sensing image forest area. Feature extraction includes two parts: adaptive color region extraction through DC (Definition Circle Model) model and corner feature extraction including suppression mechanism through edge detection model. After a series of experiments, the feature-significant extraction technique is more adaptive and accurate than other unsupervised target detection models.