Images Manipulation Technology And Physiochemical Response Of Rice Sheath Blight

Applying the image manipulation technology and vision choose the color characteristics of different rice sheath blight level, to determine the suffer injury level by percentage of lesion spot height of rice plant; correlation analysis of RLH and disease index; and the color characteristics of diseased level rice leaves and physiochemical index were tested. The results were as following:1. The image manipulation technology in rice sheath blight different diseased level situationThe leaves of rice were chosen by (2G-R-B) image, we can grained the satisfactory effect, while the whole plant was preferably segment threshod value. The bottom of paddy in the intensity histogram at 0.4, so we choose the initial value at T=0.4, 31 images in the experiment were automatismly succeed segment, except a few yellow leaves information were losed.Using image manipulation chosen the color parameters of whole plant, the results showed that: R value increasing along with the diseased level; H value put up definite downtrend. Because the color characteristics of rice exist some overlap range, the effect of division is not evidensive. So it is not feasible that chosen the color characteristics of whole plant images, and can't detect the diseased level of rice sheath blight.2. The change of color characteristics value range under different diseaesd leaves of rice sheath blightWe analysed the variety of the color characteristics of different diseased level leaves, R, G,r, S, value showed elevatory trend gradually along with the diseased level increasing, while B, b, H value were showed downtrend; B, g value change no rule.The disease level distinguishes results of various color characteristics as following:(1) Using the R value could distinguish 1 and 3 grade, 2 and 4 grade, 1 and 5 grade, the effect of disease level distinguish is best.(2) Using the R value could distinguish 2 and 3 and 4 grade.(3) Using the B value could distinguish 1 and 4 grade, 1 and 5 grade, but can't distinguish 1 , 2, 3 grade.(4) Using the b value could distinguish 2 and 3 grade. (5) Using the H value could distinguish 1 and 4 grade, 2 and 4 grade.(6) Using the S value could distinguish 2 and 4 grade.3. Application conducts distance method detecting rice sheath blight diseased levelUsing the distance method in the diseased level experiment of rice sheath blight, the results showed: correct rate was 80 percent in 2 grade, $9 percent in 3 grade, 100 percent in 4 grade, 80 percent in 5 grade. The correct rate of detective between 80%-100 %. This method can detect the diseased level of rice sheath blight.Correlation analysis of RLH of rice sheath blight and disease index, the data showed that the significant positive correlations, the correlation cocefficient was 0.8534.4. The physiological index test under rice sheath blight different disease levelThe physiological index of different level of rice sheath blight was mensurated, the results showed, the net photosynthetic rate, the SC, the TR of infected leaves were lower than CK, those index was falling along with the disease level.5. The biochemical index test under rice sheath blight different disease levelThe mensuration results of biochemical index were same as the physiological index. In the same disease level, the activity of enzyme(PPO,POD,PAL) were depressed along with the leaves position increasing; At the same leaves position, the activity of enzyme were depressed along with disease level increasing.