The Application of Improved Histogram Equalization and Image Segmentation in Satellite Image

Abstract

Reasonable image enhancement and image segmentation technologies help people to identify the object from original images, which cannot be easily recognized by naked eyes. A method of image recognition based on improved histogram equalization and image segmentation was proposed in this paper. By stretching the histogram in a custom range and excluding singular points, contrast of images was enhanced with more detail information. Then, feature vector in different color space was used to segment the object from original images. From testing, this method was proved to be efficient, simple and convenient to implement and can be used in real time system because of its simplicity.