A Cloud Image Retrieval Method Based on Shape Features

Abstract

The present invention discloses a cloud image retrieval method based on a shape feature. In effective combination with brightness temperature information of a cloud top of a cloud image, a cloud system is segmented by using an iteration threshold segmentation method; and a shape feature of the cloud system is extracted by using a geometric invariant moment that is relatively low in computational complexity and strong in robustness, thereby overcoming difficulty in analyzing the cloud image by using a conventional cloud image retrieval method based on the shape feature, and especially overcoming problems of method universality, computational complexity and robustness. In addition, according to the method, a bottom cloud image feature is transformed from euclidean space to sparse space, and a cloud image database is retrieved by using a distribution rule of the sparse space, thereby effectively alleviating a semantic gap problem of a conventional retrieval method.

Description

technical field [0001] The invention relates to a cloud image retrieval method, in particular to a shape feature-based cloud image retrieval method. Background technique [0002] Satellite cloud images can display the characteristics of various cloud systems and their evolution process from multiple angles, which is of great significance in weather monitoring, climate research or disaster relief decision-making. Each cloud system in a satellite cloud image often corresponds to different weather information. Generally, if two cloud images are similar in grayscale, shape, and texture, the weather conditions of the two cloud images at this stage have a certain degree of similarity and are of reference value. At the same time, with the development of meteorological satellite technology, each data station can receive massive cloud data on the order of gigabytes covering almost the whole world every day, and the traditional method of manually labeling cloud images has been stretch...

AI Technical Summary

Problems solved by technology

Japan has studied the deformable ellipse to extract the cloud shape characteristics of typhoon cloud system, and expressed the trajectory according to the change of ellipse position, and designed a retrieval system for typhoon cloud map: Kitamoto, but this system can only be used for typhoon cloud It is not general; Acqua et al. in Italy used point diffusion technology to characterize the shape characteristics of cloud images through position, rotation and scale, and studied the retrieval system for hurricane and non-hurricane cloud systems. This system also has application areas. Narrow problems and complex calculations; India's Deepak uses the area and perimeter of cloud clusters as shape features to implement a cloud image retrieval system, but experiments show that such features are only good for typhoon cloud systems

In China, Li Yanbing et al. use a circular "deformable model" to describe the cloud shape and extract the parameter features of cloudy blocks, but the complexity of circular extraction and matching is very large, and the calculation is time-consuming

Shangguanwei from Harbin Engineering University used the variation method to extract the shape features of cloud images, and realized a cloud image retrieval system through fuzzy similarity calculation, but the retrieval accuracy of the system is not high; Nanjing University of Aeronautics and Astronautics combined particle swarm optimization and FCM methods, First, the cloud system is clustered in the early stage to obtain the cumulus cloud system, and then the shape feature is extracted by the geometric invariant moment, and the similarity matching is performed to realize the cloud image retrieval. However, when this method is faced with complex cloud systems, it is difficult to extract comprehensive shape features. And similarity matching is not easy to achieve

Images