Development And Application Of Image Processing Technique To Diagnose Plant N Nutrition

The study was performed to explore how image technique can be used to diagnose the nitrogen deficiency in plant. The research included two parts. Firstly, the statistic methods of image histogram and arithmetic of measuring areas of 2D- materials were introduced by using the character data of plant which was in nitrogen deficiency condition, and then designed the soft with Visual C++. Secondly, soybeans were cultivated under different nitrogen level in nutrient solution; the soft was used to statistics the histogram of leaf RGB color value and to calculate the leaf area. The statistics of biomass also was done. The combination of the two ways has revealed the relationship between the digitization rule of RGB and nitrogen level in leaf.The main conclusion followed: 1.The soft has been developed under windows OS. which diagnosed the nitrogen deficiency, and itsmain function included measuring the leaf area and diagnosing the nitrogen deficiency. 2.The software of measuring the areas of 2D- materials could exactly count the leaf area. Thecorrelation coefficient of leaf area measured by traditional copy method and scanning method wassignificant high (r=0.997l). 3.The results indicated that green color value was bigger than red color value, and the blue color valuewas lower than red color value. The results were correlative with the reflectivity of light. 4.The correlation between red color value and N content in plant's leaves was negative, the same as togreen color value. The correlation between blue color value and N content in plant's leaves waspositive. The correlation between red/blue value and N content in plant's leaves was notably negative. 5. RGB color value was distinguished in different parts of leaf. In order to make sure the results wereright, we diagnosed the total part of one leaf or averaged the 9 parts of the leaf.