Study On Distribution Of Crop’s Structure In Heilongjiang

In the context of economic globalization, the food is not only the relationship of the national economy and people’s livelihood is also a kind of important strategic materials of the flow of goods in the world. Timely and accurately grasping the agricultural information, to ensure national food security, to get advantage in the international grain price has important significance. In establishing agricultural remote sensing in the operation of the system,determining points of crop planting structure space is a key ring. A lack of planting structure of the results support on the one hand, limits the accuracy of remote sensing monitoring crop condition at present, on the other hand looks like the led to the remote sensing monitoring in the division, crop model encountered great difficulties, the same problem also exists in the crop yield estimation and agricultural disaster forecast. At the same time, as a kind of basic data, crop planting structure space changing rule of crop planting structure, crop structure and the response of the global environmental change and the study of agricultural policies, formulate, adjust and aspects of the protection of national food security has very important practical significance.Crop classification recognition and extraction is the key problem in the study of the crop planting structure extraction. Because of the complexity of the crop category, remote sensing information and the existence of mixed spectral values, making just rely on the traditional classification method of spectrum of crop is often difficult to achieve ideal results. Because the crop has the characteristics of the seasonal rhythmicity and phenological change regularity, and the same kind of crop in the same area with relatively sTab. growth. When using the time series of remote sensing data of phase change rule can be used as the basis for identification of different crop types.In this paper, on the basis of analysis of NDVI time-series data, we realized the digital expression of phenological information. In this paper, through the TIMESAT software, on the basis of the processing of NDVI curve smoothly, to extract the11phenological key values, used to describe the curve shape and the difference between crop phenological period. On this basis, the study on extraction of cultivated land area in Heilongjiang province. Through the statistical analysis of different land types phenological key values, to establish the classification system and classification threshold, the use of object-oriented hierarchical classification based on rules, land use type of Heilongjiang province, to extract the cultivated land area. Heilongjiang province four to span is big, the entire study area light temperature have significant differences, different areas of the same crop phenological period. In order to achieve the extraction of crop planting structure, this paper has carried on the partition of cultivated land, keep the area relatively uniform phenological parameters. Based on partition, this paper USES the object-oriented method, the selection of the main components of the phenological key values, one by one in the division of the classification of the area for planting pattern extraction, end up with Heilongjiang province planting pattern information. In this paper, we got the following research conclusions:1. In this paper, the TIMESAT provides three methods for filtering the repeated experiment and contrast analysis, it is concluded that Savitsky Golay filtering method of minimum root mean square error, so the original curve fitting degree is the best. At this point because the other two methods.2. Using fixed threshold and dynamic threshold method to extract the information phenological key values, compared with Chinese lunar calendar information found that the fixed threshold method in SGS and EGS point of the theoretical results fit well with the Chinese lunar calendar information, is superior to the dynamic threshold method; Extract the NDVI curve described in11phenological key values in the form of information.3. The comparison and analysis of various land types in Heilongjiang province key value differences in parameters, sorting out the key threshold to distinguish the various land types and classification system, using object-oriented hierarchical classification based on rules to classify land use type. Not reached the classification accuracy, extract the scope of cultivated land. But also proved that the phenological information can be used for the classification of land use, the object-oriented method can be realized and phenological information utilization.5. The partition of crop planting structure in Heilongjiang province are extracted. In cultivated land division, on the basis of the agricultural insurance companies provide insurance block samples, using the object-oriented method of regional land for segmentation, the optimal scale selection operation, finally get the crop pattern of Heilongjiang province.6. Crop structure of Heilongjiang province was analyzed, and the statistical pattern of various types of planting area, the three main crops of soybean, corn, rice planting spatial pattern characteristics are analyzed. Summarizes the distribution law of three kinds of crops in Heilongjiang province for the partition with macro scope, for the future agricultural disaster assessment, provide basic information yield prediction, etc.This study is based on phenology information data has realized the extraction of crop planting structure, has made certain research progress, to provide basic data for the agricultural remote sensing monitoring.