Research On Earthquake Prediction Algorithm Of Infrared Cloud Images Based On Image Processing

The number of global natural disasters has gradually increased, and nearly 200 million people suffer from natural disasters every year. There are many kinds of natural disasters in our country, and the distribution is not uniform. More than 70% of the population, more than 80% of the industry and agriculture and cities are threatened by a variety of natural disasters. So the prediction of the earthquake is particularly important to avoid more loss of life and property. Many achievements in the field of predict earthquake from remote sensing images have been obtained in the last few decades. But the traditional prediction methods presented do have the limitations of can't forecast epicenter location accurately and automatically. Thus, a new predicting earthquakes method through the identification of thermal infrared anomaly cloud and continued tracking the position and frequency of abnormal clouds is proposed in this paper.According to the theory and method of seismic prediction of thermal infrared anomaly cloud, the steps of the algorithm are given in detail. The method is mainly divided into two parts: the identification and tracking of thermal infrared anomaly cloud.There are two steps: sample training and classification in the recognition part. First training samples, accurate abnormal cloud is chosen as a positive sample, non-anomalous cloud is chosen as a negative sample and calculate the texture feature vector of them. The texture feature vector as the input, whether it is the exception of the cloud as the output,the neural network classifier trained to get abnormal cloud.Next is the classification and recognition of each input image. First step is image preprocessing, filter out the region which is not likely to be abnormal and strengthen the suspicious areas in the infrared cloud images. Then calculate the texture feature vector of cloud around each point, and use them as the input of the classifier to classify each pixel.Clustering and filtering the classification results to extract the suspected abnormal cloud region.For the tracking part, by tracking the thermal infrared image of a period of time, if a regional anomaly cloud reproduction of high frequency, it is considered that the position of a possible earthquake. And then to estimate the location of the earthquake epicenteraccording to the evolution of abnormal position cloud CenterIn this paper, the experimental results of seismic inversion are proved that Thermal infrared anomaly cloud recognition algorithm can effectively realize automatic earthquake prediction. This method not only can accurately predict the earthquake center, but also has a certain prediction effect on the magnitude and time of earthquake.