Abstract:

A cloud detection algorithm combining ACCA (automatic cloud cover assessment) with WSVM (weighted support vector machine) is proposed to solve the problem that ACCA algorithm is difficult to detect the translucent cloud on Landsat images. Firstly, the ACCA algorithm is used to divide image pixels into cloud pixels, non-cloud pixels and undetermined pixels based on the atmospheric radiation characteristics of cloud in different bands and the spectral characteristics of Landsat ETM + image data. Then using the spectral properties of cloud to construct feature vectors, and using WSVM algorithm to detect the undetermined pixels, the cloud detection results of all the images are obtained. Experimental results show that this method not only has the advantages of ACCA cloud detection algorithm, but also has good detection effect for the translucent cloud which is hardly identified by ACCA.

Key words: Landsat images, cloud detection, ACCA, WSVM