| As great development of earth observation techniques, more and more remote sensing data with different spatial resolutions, spectral bands and working modes are producing day by day. These progresses, in the one hand, make a good prospect for remote sensing applications; in the other hand, bring a serious problem for information extraction techniques. The problem is how to draw useful information needed by a user from huge number of remote sensing data more correctly and efficiently. Obviously, it is a big barrier not only for progresses of remote sensing techniques, but also for remote sensing applications. Therefore, development of Information Extraction Techniques (IET) from remote sensing data and with high efficiency and good quality becomes an urgent, significant, critical and difficult task for scientists and technicians in the field of earth observation. Based on author's review of IET development both in domestic and oversea, and the best practices conducted by author and her colleagues in the institute, and author's Ph. D thesis research during past three years, a grouping interpretation technique for remote sensing images is presented. It is one of the ways to solve the problem as mentioned above. The research on the grouping interpretation technique for remote sensing images consists of three parts: study on principle and methodology of grouping interpretation, development of Grouping Interpretation System (GrIS) and applications of the technique and its methods to some typical areas and tasks. The main results of the thesis study are presented as following:(1) The principle and methodology for grouping interpretation technique of remote sensing images is systematically presented in the thesis. They have dealt with a set of basic problems for the technique, such as its basic conception, imagery elements, interpretation keys, operation procedures and both assessments of the interpretation with accuracy and efficiency. Among them, classification and establishing procedure of interpretation keys with three level structures, and the model for assessment of interpretation efficiency and its calculation method are innovated.a(2) A Grouping Interpretation System for Remote Sensing Images (GrIS) was developed based on application task requirements, visual interpreting procedure and manner, and multi-technique integration. GrIS can operate in both single-computer mode and multi-computer mode with client/server structure in LAN and WAN environment. Its functions with three level structures consist of module functions, system functions and group system functions and can flexibly and easily meet requirements from different users and tasks. The system can be used to deal with remote sensing inventory and mapping, quality control of interpretation task, detection and recognition of important targets and image interpretation training, etc. The task management functions of the system can easily deal with task assignment, progress monitor, and dispatch of resources, assessment of accuracy and efficiency, and integration of results from different places, etc. Therefore, GrIS is a system quite different from any existing system in the field, including Image Interactive Interpretation System (13S) involved by author before.(3)Applications of grouping interpretation technique and its methods in typical fields was conducted in the pilot studies. The fields include monitoring of natural disasters, land use/land cover mapping, ecological and environment monitoring, and technical training, etc. Among them, the establishing procedure of interpretation keys for urban land use mapping in Tangshan City has been paid much more attention to. Those pilot applications as mentioned above demonstrated the advantages and potential of grouping interpretation technique, also strongly supported the development of the technique and methods.For the future, GrIS will be improved from some aspects of functions and performances. Obviously, making GrIS more intelligent and more automated is one of more significant matters.
|
PDF Full Text Request