An efficient algorithm for automatic road extraction is presented based on fuzzy mask in this paper. The fuzzy mask concepts are employed in order to perform road pixels extraction. The aim of the proposed approach is to attribute a road membership value to each pixel. Firstly, a set of fuzzy mask sets, which represent basic road structures, is defined. The road structure, defined as a logical concatenation of these elementary predefined 2D structures, is used in order to enhance the road membership values. The input data of the SPOT panchromatic band of remote sensing is required, and it is applied in the system. Obtained experimental results show that this method is efficient in extracting and defining road networks from satellite and aerial imagery.

Detection and extraction of the dim moving small objects in the infrared image sequences is an interesting research area. A system for detection of the dim moving small targets in the IR image sequences is presented, and a new algorithm having high performance for extracting moving small targets in infrared image sequences containing cloud clutter is proposed in the paper. This method can get the better detection precision than some other methods, and two independent units can realize the calculative process. The novelty of the algorithm is that it uses adaptive thresholding technique of the moving small targets in both the spatial domain and temporal domain. The results of experiment show that the algorithm we presented has high ratio of detection precision.

With an increasing number of images collected every day from different sensors, automated registration of multisensor/multi-spectral images has become an important issue. Image registration is a very important precondition of image fusion. A wide range of registration techniques exists for different types of applications and data sources, however no algorithm is known that can accurately register multi-source images consistently. A new method of remote sensing images registration is presented in this paper. The development of this new automatic image registration method is based on Fourier transform and mutual information, which use wavelet decomposition technique. The result of experiment shows that this algorithm can effectively solve the image registration of remote sensing images with quite different diversities, and it has a high value in use.