Design And Implementation Of Rice Extraction And Parameters Retrieval System Based On Microwave Remote Sensing

Rice is a staple food for over 60% of population in China. Rice production is directly related to the food security of this country. Remote sensing technology has advantages such as macroscale, comprehensiveness, real-time and dynamics, and is playing an important role in paddy rice field extraction, growth monitoring and yield prediction. In comparison with optical remote sensing, due to its working on longer wavelength, the synthetic aperture radar (SAR) is less influenced by clouds and rain, and can penetrate some media on surface, and has the capabilities of all-time, all-weather observations. In recent years, SAR applications in agriculture have been conducted widely. Research on rice extraction and parameters retrieval can provide direct or indirect key variables of rice growth and yield on regional scale. And also, it can provide scientific guidance on agricultural production and provide decision-making basis for macroeconomic regulation and control of grain, therefore it has important application value.However, the application system based on microwave remote sensing is still unavailable at present. Development of rice application system based on microwave remote sensing and construction of an interface visualization application can provide a powerful tool for rice filed discrimination, simulation of rice scattering characteristics, and retrieval of rice canopy parameters. This system can also provide a quick and reliable method for fast prediction of rice yield at regional scale. MATLAB has advantages in scientific computing and visualization with a useful toolbox. And also, it can be used to create user interface and call other languages program coding with C, C++ and Fortran, etc. Therefore, it’s very suitable for developing microwave remote sensing application system.This paper mainly introduced some key techniques for the rice extraction and parameters retrieval system based on microwave remote sensing at the MATLAB environment, including its handle graphics object, GUI incident response and data transmission, hybrid programming with Fortran and Access. The function and structure of this system including several module such as sampling data management, rice extraction, RCSM simulation, sensitivity analysis, parameter retrieval and auxiliary tools. Furthermore, this paper discussed the key problems in extraction of rice area, simulation of scattering characteristics, retrieval of canopy parameter, and design of specific algorithm and process. By running the instance data observed in a case study area, it was proved that all modules of this system is able to work well in functionability of corresponding task with stable and safe performance.