Constructional Fixed Asset Investment Trend Monitoring Based On Multi-source Remote Sensing Data

Fixed assets investment is crucial driver for landscape pattern change and the development of city economy. Fixed asset investment trend monitoring is a hot issue in recent years. This paper took Zhongmou and Beibei as study areas. Starting with the research of relationships between city expansion and fixed asset investment,methodology of fixed asset investment feature extraction, driving forces of landscape pattern, this paper proposed a series of monitoring methods in different scales. The major contents and conclusions of this research are as follows:(1) By modifying the panel data of China region fixed assets investment from1990 to 2010 and extracting city built-up areas from Remote Sensing data in corresponding years, regression correlation had been analyzed which show that the siphon effect of fixed assets investment is existing simultaneously at large scale. It is feasible to monitor the fixed assets investment trend by RS technology. The data reveals that with the gradual extent of China’s reform, the investment scale and the expansion area relationship turns from coupling to decoupling. Moreover, the driving force weight of the investment is decreased. Unit expansion area of fixed assets investment indicator suggest that the intensive effect of Chinese cities’ expansion is remarkable.(2) The algorithms for fixed assets investment features detection are proposed.The experimental results show that this new algorithm can be used to detect constructions and remove disturbing soil robustly when using the VHR Remote Sensing data and high-resolution data. Also, it provides a more portable results comparing train a random forest classifier with traditional ones and can be used to detect other more feature targets. In addition, as tower crane is the particular construction equipment for investment projects, the algorithm combining mathematical morphology and geometry algorithm can detect the target precisely both in positioning and quantity with experimental verification.(3) The investment project classification system of fixed assets investment monitoring includes 15 categories is established. The classification characteristicindex BBI and IPBI are constructed, with processing by object-oriented algorithm and analyzing a variety of optical and texture characteristic value, Choi steel constructions and other temporary buildings target objects are detected, we got the access to the microscopic scale investment hotspot distributed effectively.(4) The evolution of landscape pattern is studied through dynamic monitoring.By analyzing the landscape pattern in nearly 15 years with four time plots which are2000, 2005, 2010 and 2015. The framework of landscape indices on plains and mountainous areas are developed and analyzed separately.