Diagnosis Of N Status Of Winter Wheat Using Digital Image Processing And Soil-Plant Testing Techniques

The most widely used methods to assess the nitrogen (N) status of winter wheat (Triticum aestivum L.) are the determination of plant total N by combustion, the testing of nitrate in the stem tissue and the use of SPAD readings. However, due to their labor requirements or high costs these methods can hardly be applied to the huge wheat growing areas of the Northern China Plain. In this study, digital image analysis method, which combined with soil and plant N rapid test, was studied for N status diagnosis and N fertilization recommendation. In this thesis, the experiment was conducted in Beijing suburb within a multi-factorial field trial with three levels of N. The possibility of using digital image analysis method for plant N status diagnosis was studied, and the standard method of digital image acquisition and saving was established. In addition, aerial true color and infra-red images analysis method, which combined with soil and plant N rapid test, was studied for N recommendation in a large scale. The main results are as follows:The digital image processing technique could be used of N status detection of winter wheat. The results show significant inverse relationships between greenness intensity and canopy total N, SPAD readings, plant nitrate concentration and soil N supply at shooting and flowering stage. Compared with other three green index of plant canopy, the normalized greenness intensity G/(R+G+B) was much better because of its better correlations with canopy total N, SPAD readings, plant nitrate concentration and soil N supply.For taking digital image of wheat canopy in field, the angle between digital camera and ground could be 30~60?. JPEG format with a pixel resolution of 1024X768 could be used for digital images saving. The brightness of the field background and the light condition did not affect the image acquisition and the results of colour analysis. Compared with digitized pictures using optic camera and through scanning, the method obtaining image by digital camera directly could get better quality image, and more convenient.Field digital image analysis method combined with soil and plant rapid test could be used for N fertilization recommendation. In this field experiment, when G/(R+G+B) was less than 0.408, winter wheat was in N surplus state, did not need N topdressing at shooting stage; when G/(R+G+B) was 0.408-0.422, winter wheat was in optimum N supply state; and when G/(R+G+B) exceeded 0.422, winter wheat was in N stress state at shooting stage.Aerial true color image combined with soil and plant rapid test could be used for N fertilization recommendation at large scale field. The normalized redness intensity of R/(R+G+B) could be a suitableindicator for N status diagnosis at shooting stage. When R/(R+G+B) was less than 0.256 winter wheat was in N surplus state; when R/(R+G+B) was 0.256-0.334, winter wheat was in optimum N supply state; and when R/(R+G+B) exceed 0.334, the winter wheat was in N stress state.The plant canopy indexes of RVI, NDVI and GNDVI, which obtained from aerial infra red images analysis, were good correlated with plant total N content, SPAD readings and sap nitrate concentration. When NDVI was exceed 0.269, winter wheat was in N surplus state, did not need N topdressing at shooting stage; when NDVI was 0.172~0.269, winter wheat was in optimum N supply state at shooting stage; and when NDVI was less than 0.172, winter wheat was in N stress state at shooting stage.