Research Of The Methods In Removing Cloud Of Remote Sensing Image

With the increasingly widespread application of remote sensing data, quality of the images has become the focus of researchers. Remote sensing imaging process affected by many factors, cloud is one of the most common confounding factors. We can remove thin cloud of single image to enhance surface features information from the perspective of image enhancement.This article discusses remote sensing image degradation model, image data calibration method, image noise reduction and image enhancement technology. First, analyze remote sensing imaging factors, imaging mechanism of thin cloud, thin cloud regional characteristics; we study and simulate the homomorphism filtering, mathematical morphology and wavelet algorithm. Then study the spatial and frequency domain image enhancement techniques, image enhancement technology, and its combination with the existing algorithms to achieve the thin cloud removal algorithm. The improved algorithm combines the frequency domain and spatial domain, first select the appropriate structural elements and characteristics of good Butterworth filter, through morphological filtering and homomorphism filter to remove a lot of thin cloud; then expand contrast of the results image, and enhanced image details by unsharp mask filter. Simulation and a large number of experiments show that the improved algorithm can reserve the maximum surface information while maximizing the removal of thin cloud to get good effect.Finally, analyze human visual properties to make a subjective assessment. Analysis and comparison of algorithms by statistical parameters shows that the improved algorithm is feasible and effective. Design system which can easily select the image, algorithm and determine the appropriate parameters to quickly achieve thin cloud removal, visual indication of the effect of each method.