Monitoring Test Based On The Beijing-1 Small Satellite Data Macro Land Use

This research chooses Tianjin, Jiangsu Yixingshi as well as Inner Mongolian arun banner three regions is the research area, analyzes Beijing-1 Micro-satellite Data multi-spectrum data in macroscopic land utilization monitor application ability, and constructs the land utilization classification indicator system. In carries on experimental to the Experimental area in the foundation, according to the Experimental area characteristic, the union Experimental area's natural economy factor, chooses the non-surveillance classification, the surveillance classification, the lamination classification as well as the man-machine interaction interprets four classification approaches, carries on the land utilization classification, and carries on the precision appraisal to the classified result, the definite Experimental area classifies the indicator system finally; Carry on the Contrastive analysis to the Experimental area each kind of classification approach's accuracy and the efficiency, determined that suits the pilot area land utilization information extraction the best classification approach. Causes Beijing-1 Micro-satellite Data better to manifest in macroscopic land utilization monitor ability.This article main conclusion is as follows:1. Beijing-1 Micro-satellite multi-spectrum data image quality is good, the spectrum information is rich, the texture structure is clear, may the clear area divide the cultivated land, the garden, the lawn, the construction land, the waters wetland as well as other lands from the image and so on first-level land classification, to in land classification's second-level classified paddy field, the arid land, the forest land and so on can also the good extraction.2. Uses the non-surveillance classification, the surveillance classification, the lamination classification as well as methods and so on man-machine interaction interpretation carries on the classification to Beijing-1 Micro-satellite multi-spectrum image, obtained the good classified result. The Tianjin land classification non-surveillance classifies, the surveillance classificationas well as the lamination classification overall precision respectively is 61.14%, 62.75%, 91.87%; The Yixing land classification non-surveillance classifies, the surveillance to classify, the lamination classification as well as the man-machine interaction interpretation overall precision respectively is 63.25%, 74.33%, 89.58%, 92.33%; The arun banner land classification non-surveillance classifies, the surveillance to classify, the lamination classification as well as the man-machine interaction interpretation overall precision respectively is 55.0%, 68.2%, 82.79%, 88.33%. The synthesis compares each classification approach the land utilization classification precision and the classified efficiency, three pilot area's best classification approaches were the lamination classification approach, the classified precision have achieved above 80%. Has achieved the classification recognition capability to the second-level classification, obtained the ideal classified result, explained that Beijing-1 Micro-satellite multi-spectrum data has the usability in the macroscopic land utilization monitor research area, may and so on in the related work apply in the national land macroscopic monitor investigation.3. As a result of this research institute application's multi-spectrum data's phase's limit, Beijing-1 Micro-satellite data has not been able to dig completely in land utilization research area's many potentials according to comes out, but also waits for in the application further develops after near, further studies Beijing-1 Micro-satellite data in complex terrain landform region land utilization information extraction ability.