“A research team has proposed an innovative unsupervised superpixel level change detection method to address the long-standing challenges in the field of change detection, such as inaccurate detection results caused by noise interference and the influence of special terrain. This method significantly improves the accuracy and stability of change detection by combining canonical correlation analysis and histogram regularization. The team first preprocessed and superpixel segmented two remote sensing images at different time points, and then calculated weights based on superpixel scale and unchanged probability. Next, through superpixel level multivariate change detection and histogram regularization, the change features are accurately extracted. Finally, using weighted images, classical methods, and change features for decision fusion, accurate change detection result maps are obtained. To verify the effectiveness of the method, the research team conducted experiments on three hyperspectral test datasets and one multispectral test dataset. The results showed that the method performed the best in both OA and Kappa metrics on the four test datasets, with OA reaching over 90%. Compared with the highest accuracy among other methods, the OA of this method has improved by 4.41%, 3.44%, 1.74%, and 0.19%. This research achievement provides a new solution for the field of change detection, not only improving detection accuracy, but also expanding the application scope of remote sensing image analysis. In the future, this method is expected to play an important role in fields such as environmental protection, urban planning, and disaster monitoring.”