Cloud removal of remote sensing image based on multi-output support vector regression

Removal of cloud cover on the satellite remote sensing image can effectively improve the availability of remote sensing images. For thin cloud cover, support vector value contourlet transform is used to achieve multi-scale decomposition of the area of thin cloud cover on remote sensing images. Through enhancing coefficients of high frequency and suppressing coefficients of low frequency, the thin cloud is removed. For thick cloud cover, if the areas of thick cloud cover on multi-source or multi-temporal remote sensing images do not overlap, the multi-output support vector regression learning method is used to remove this kind of thick clouds. If the thick cloud cover areas overlap, by using the multi-output learning of the surrounding areas to predict the surface features of the overlapped thick cloud cover areas, this kind of thick cloud is removed. Experimental results show that the proposed cloud removal method can effectively solve the problems of the cloud overlapping and radiation difference among multi-source images. The cloud removal image is clear and smooth.

CLOUD IMAGE DETECTION BASED ON MARKOV RANDOM FIELD

In order to overcome the disadvantages of low accuracy rate, high complexity and poor robustness to image noise in many traditional algorithms of cloud image detection, this paper proposed a novel algorithm on the basis of Markov Random Field (MRF) modeling. This paper first defined algorithm model and derived the core factors affecting the performance of the algorithm, and then, the solving of this algorithm was obtained by the use of Belief Propagation (BP) algorithm and Iterated Conditional Modes (ICM) algorithm. Finally, experiments indicate that this algorithm for the cloud image detection has higher average accuracy rate which is about 98.76% and the average result can also reach 96.92% for different type of image noise.

A Method of Identifying Thunderstorm Clouds in Satellite Cloud Image Based on Clustering

In this paper,the clustering analysis is applied to the satellite image segmentation,and a cloud-based thunderstorm cloud recognition method is proposed in combination with the strong cloud computing power.The method firstly adopts the fuzzy C-means clustering(FCM)to obtain the satellite cloud image segmentation.Secondly,in the cloud image,we dispose the‘high-density connected’pixels in the same cloud clusters and the‘low-density connected’pixels in different cloud clusters.Therefore,we apply the DBSCAN algorithm to the cloud image obtained in the first step to realize cloud cluster knowledge.Finally,using the method of spectral threshold recognition and texture feature recognition in the steps of cloud clusters,thunderstorm cloud clusters are quickly and accurately identified.The experimental results show that cluster analysis has high research and application value in the segmentation processing of meteorological satellite cloud images.

Estimation of Cloud Motion Using Cross-Correlation

This paper describes the estimation of cloud motion using lag cross-correlation. In order to compute the lag cross correlation, the Bayes Decision method is used first to identify cloud and surface of earth. Then cloud motion vectors are retrieved at a subset of points through multiple applications of a cross-correlation analysis. An objective analysis is used to define displacement at every satellite pixel throughout the domain and smooth the local inconsistencies. Cloud motions are then produced with a backward trajectory technique by using these displacement vectors.

Ground-Based Cloud Using Exponential Entropy/Exponential Gray Entropy and UPSO

Objective and accurate classification model or method of cloud image is a prerequisite for accurate weather monitoring and forecast.Thus safety of aircraft taking off and landing and air flight can be guaranteed.Thresholding is a kind of simple and effective method of cloud classification.It can realize automated ground-based cloud detection and cloudage observation.The existing segmentation methods based on fixed threshold and single threshold cannot achieve good segmentation effect.Thus it is difficult to obtain the accurate result of cloud detection and cloudage observation.In view of the above-mentioned problems,multi-thresholding methods of ground-based cloud based on exponential entropy/exponential gray entropy and uniform searching particle swarm optimization(UPSO)are proposed.Exponential entropy and exponential gray entropy make up for the defects of undefined value and zero value in Shannon entropy.In addition,exponential gray entropy reflects the relative uniformity of gray levels within the cloud cluster and background cluster.Cloud regions and background regions of different gray level ranges can be distinguished more precisely using the multi-thresholding strategy.In order to reduce computational complexity of original exhaustive algorithm for multi-threshold selection,the UPSO algorithm is adopted.It can find the optimal thresholds quickly and accurately.As a result,the real-time processing of segmentation of groundbased cloud image can be realized.The experimental results show that,in comparison with the existing groundbased cloud image segmentation methods and multi-thresholding method based on maximum Shannon entropy,the proposed methods can extract the boundary shape,textures and details feature of cloud more clearly.Therefore,the accuracies of cloudage detection and morphology classification for ground-based cloud are both improved.

Tropical Cyclone Cloud Image Segmentation by the B-Spline Histogram with Multi-Scale Transforms

An efficient tropical cyclone（TC） cloud image segmentation method is proposed by combining the curvelet transform,the cubic B-Spline curve,and the continuous wavelet transform.In order to enhance the global and local contrast of the original TC cloud image,a second-generation discrete curvelet transform is implemented for the original TC cloud image.Based on our prior work,the low frequency components are enhanced by using an incomplete Beta transform and the genetic algorithm in the curvelet domain. Then the enhanced TC cloud image is used to segment the main body of the TC from the TC cloud image. First,pre-processing is implemented by B-Spline curves to the original TC cloud image to remove unrelated small cloud masses.A region of interest（ROI） which includes the main body of TC can thus be obtained. Second,the gray-level histogram of ROI is obtained.In order to reduce oscillations of the histogram,the gray-level histogram is smoothed by cubic B-Spline curves and the B-Spline histogram is obtained.The one dimensional continuous wavelet transform is employed for the curvature curve of the B-Spline histogram. A new segmentation cost criterion is given by combining threshold,error,and structure similarity.The optimally segmented image can be obtained by the criterion in the continuous wavelet domain.The optimally segmented image is post-processed to obtain the final segmented TC image.The experimental results show that the main body of TC can be effectively segmented from the complex background in the TC cloud image by the proposed algorithm.

A CONVECTIVE RAINFALL ESTIMATE TECHNIQUE DEVELOPED BY NSMC OF CHINA

In order to reveal the relation between strong convective cloud characteristics and rainfall rate, over 20000 hourly raingauge data from 333 weather stations and the corresponding 4000 convective cloud infrared images of GMS-4 during the period of 1992—1994 in Henan,Hubei and Sichuan Provinces were studied.The results show that cloud top temperature,temperature gradient,the growth of cloud,overshooting top and the normalized distance between a cloud covering pixels and the cluster center have certain relations to cloud precipitation.These relations can vary with different geographical regions.Based on the study above,a convective rainfall estimation technique was developed by the scientists in National Satellite Meteorological Center of China.Its average error is 30% for daily precipitation with a correlation coefficient of 0.69.