 This paper proposes a novel weighted semantic segmentation algorithm for remote sensing images. This algorithm uses effective sample theory to calculate weights for each pixel in the image, which allows for more accurate segmentation results when compared to other algorithms. The algorithm was tested on two datasets, one for land use and land cover classification and another for forest fire burning area classification. Results showed that the proposed algorithm outperformed other algorithms in terms of both accuracy and precision. This article was authored by Zheng Zhou, Chen Sheng, Xiaodong Lu, and others.