TY - GEN
T1 - Diagnosis of nitrogen nutrition of rice based on image processing of visible light
AU - Yuan, Yuan
AU - Chen, Lei
AU - Li, Miao
AU - Wu, Na
AU - Wan, Li
AU - Wang, Shimei
N1 - Publisher Copyright:
© 2016 IEEE.
PY - 2017/1/13
Y1 - 2017/1/13
N2 - A field experiment was carried out to explore the feasibility of visible image processing technique in rice nitrogen nondestructive diagnosis. Based on the image processing technology of visible light, this study investigated the relationship between the nitrogen status of rice and the greenness of rice leaves, which are captured by digital camera. Data were collected on Luhan 1 rice variety and included photos and SPAD values of leaves. Linear regression s were performed between the extracted color feature parameters and the distribution characteristics of rice leaf SPAD values at different positions (leaf tip, leaf middle and leaf base), in order to identify the best color feature parameters, the best positions and the optimum period for rice nutrition diagnosis. The results showed good linear correlation between the SPAD values and some color features, including NRI, Hue, and DGCI in the whole growth period, where the determination coefficients were 0.92<, 0.99< and 0.99< respectively. The experimental results demonstrated that leaf middle was a more suitable position for rice nitrogen diagnosis than leaf tip and leaf base. Both the tillering stage and booting stage of rice are the crucial diagnosis stages for assessing the rice nitrogen status. The experimental results indicated that combining SPAD and visible image processing technique is usable in diagnosing the nitrogen nutrition of rice. The proposed method is fast, non-destructive and easy to apply.
AB - A field experiment was carried out to explore the feasibility of visible image processing technique in rice nitrogen nondestructive diagnosis. Based on the image processing technology of visible light, this study investigated the relationship between the nitrogen status of rice and the greenness of rice leaves, which are captured by digital camera. Data were collected on Luhan 1 rice variety and included photos and SPAD values of leaves. Linear regression s were performed between the extracted color feature parameters and the distribution characteristics of rice leaf SPAD values at different positions (leaf tip, leaf middle and leaf base), in order to identify the best color feature parameters, the best positions and the optimum period for rice nutrition diagnosis. The results showed good linear correlation between the SPAD values and some color features, including NRI, Hue, and DGCI in the whole growth period, where the determination coefficients were 0.92<, 0.99< and 0.99< respectively. The experimental results demonstrated that leaf middle was a more suitable position for rice nitrogen diagnosis than leaf tip and leaf base. Both the tillering stage and booting stage of rice are the crucial diagnosis stages for assessing the rice nitrogen status. The experimental results indicated that combining SPAD and visible image processing technique is usable in diagnosing the nitrogen nutrition of rice. The proposed method is fast, non-destructive and easy to apply.
KW - digital image
KW - Nitrogen diagnosis
KW - rice
KW - SPAD
KW - visible light
UR - http://www.scopus.com/inward/record.url?scp=85013830562&partnerID=8YFLogxK
U2 - 10.1109/FSPMA.2016.7818311
DO - 10.1109/FSPMA.2016.7818311
M3 - 会议稿件
AN - SCOPUS:85013830562
T3 - Proceedings - 2016 IEEE International Conference on Functional-Structural Plant Growth Modeling, Simulation, Visualization and Applications, FSPMA 2016
SP - 228
EP - 232
BT - Proceedings - 2016 IEEE International Conference on Functional-Structural Plant Growth Modeling, Simulation, Visualization and Applications, FSPMA 2016
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2016 IEEE International Conference on Functional-Structural Plant Growth Modeling, Simulation, Visualization and Applications, FSPMA 2016
Y2 - 7 November 2016 through 11 November 2016
ER -