Leaf chlorophyll concentration using random forest
Download as PPTX, PDF1 like1,118 views
This document discusses using imaging spectroscopy to estimate chlorophyll content in soybean leaves. It begins with an introduction to near infrared spectroscopy and its history. It then describes how a field imaging spectroscopy system was used to collect spectral data from soybean leaves. Random forests regression and the PROSPECT radiative transfer model were used to establish a model for estimating chlorophyll concentration from the spectral data. The model was able to accurately estimate chlorophyll content in soybean leaves and has potential applications for precision agriculture management and monitoring crop health at larger scales using remote sensing.
Leaf chlorophyll concentration using random forest
- 2. 2 INTRODUCTION HISTORY MATERIALS AND METHODS FIELD IMAGE SPECTROSCOPY RESULTS AND DISCUSSION IMAGING SPECTROSCOPY APPLICATIONS CONCLUSION AND FUTURE WORK REFERENCE
- 3. NIR spectroscopy is a vibrational spectroscopic method belongs to the infrared light spectrum which is very close to the visible region (from about 750 to 2500 nm). 3 Figure 1.1 Range of Electromagnetic Radiation
- 4. NIR is a spectroscopic method based on molecular overtones and combination of C-H ,O-H,N-H ions. There are two laws which govern the basics of vibrational spectroscopy Hooke’s law and Frank principle 4 Figure 1.2: Vibrational transition
- 5. Chlorophyll Estimation using NIR Chlorophyll is an extremely important biomolecule, critical in photosynthesis, which allows plants to absorb energy from light. The chlorophyll content in leaf is an indicator of grow health condition and the soybean yield. Imaging spectroscopy has great potential for estimating chlorophyll content of soybean dynamic, rapidly. The objective of this presentation is to combine random forests and field imaging spectroscopy for estimating chlorophyll content in soybean at leaf scale. 5
- 6. Three centuries ago Sir Isaac Newton published in his ‘Treatise of Light’ the concept of dispersion of light. The history of NIR is begins with William Herschel in 18th century. The term spectroscopy was first used in the late 19th century and provided the empirical foundations for atomic and molecular physics. A large number of spectral indices have been developed for the study of chlorophyll content based on leaf reflectance. Reviews of different spectral indices developed for estimating chlorophyll content are offered by Bannari et al. (2007), He et al. (2006) and Haboudane et al. (2008). 6
- 7. 7
- 8. 8 A. Study site Field spectroscopy was carried out with an ASD FieldSpec3 in summer 2009, at the farmlands of city of Chang’chun, Jinlin province. Chlorophyll content of soybean was taken as researching target. The spectral range between 350-1250 nm was used for the retrieval of leaf chlorophyll concentration. Leaf chlorophyll concentration in soybean was measured by SPAD502. SPAD 502 Plus Chlorophyll Meter
- 9. 9 B. Field campaign data Field campaigns in the soybean field were carried out under clear sky conditions. Reflectance of soybean was collected by ASD3 spectrometer, with a 350-2500nm spectral range. C. PROSPECT model PROSEPCT is a radiative transfer model to simulate the leaf reflectance spectra with the range of 400 to 2500 nm. And it only needs a small amount of input parameters like chlorophyll content, moisture, dry matter content and structure parameter.
- 10. Random forests is a classification and regression algorithm originally designed for statistical machine learning method, which is created by Breiman. Random forests were adopted to train the training data set, in order to establish leaf chlorophyll content estimation model. Advantages: Non parametric, Non sensitive to overfitting, Calculate classification error. 10
- 11. 11 FIELD IMAGE SPECTROSCOPY A field imaging spectrometer system was designed for agriculture applications. FISS was used to gather spectral information from soybean leaves. Leaf estimation model of chlorophyll concentration was applied to the validation data set to estimate leaf chlorophyll content of soybean in the research area and validation data set was established based on proximal hyperspectral data.
- 12. FISS CONSISTS OF Opto-mechanical system: Scanning mirror, Optical lenses, Spectroscopic devices and a Charge-coupled device (CCD) camera. Electronic system: Power and Motor control circuits computer system: hardware and software Hardware is a portable laptop computer, and the software includes the FISS operating software, data acquisition software and data processing software. 12
- 13. 13 Actual photograph of FISS field measurements.
- 14. Linear regression (LR) Stepwise multiple linear regression (SMLR) Partial least squares regression (PLSR) Support-vector machine (SVM) regression 14
- 15. 15 STATISTICAL DESCRIPTION OF SOYBEAN AT SAMPLING SITES
- 16. 16 Effect of number of trees in RF-PROSPECT model for estimating chlorophyll content.
- 17. 17 Validation of the soybean chlorophyll content estimation model.
- 18. • Fast • Simple sample preparation • Non destructive • Online for process applications • Need for calibration • Opportunity for data analysis
- 19. Overlapping bands (combination), not easy to interpret. Usually not for trace level analysis.
- 20. 20 APPLICATIONS Forecasting corn yield with imaging spectroscopy. Image spectroscopy to mapping canopy concentration in the nitrogen. Near Infrared Spectroscopic Analysis in the Food Industry and Research. Image spectroscopy in Forest application.
- 21. Other Application are Imaging spectroscopy for scene analysis. Mineral mapping and Application. Medical field Industrial uses Astronomical spectroscopy 21
- 22. CONCLUSION 22 This work established the soybean chlorophyll based on random forests and PROSPECT, The estimation model can be used as an effective tool for estimation of soybean chlorophyll content, and can be adopted to precision agriculture management. Future study will concentrated on scaled up the field estimation model to satellite remote sensing level, which will monitor the soybean’s health condition in a large scale.
- 23. 2. D.Sathis Kumar et al“ Near Infra Red Spectroscopy- An Overview” International Journal of ChemTech Research CODEN( USA): IJCRGG ISSN : 0974-4290 Vol. 3, No.2, pp 825-836, April-June 2011. 3.M.E. Schaepman, S. L. Ustin, and A. J. Plaza, “Earth system science related imaging spectroscopy—An assessment, ”. Remote Sensing of Environment, vol. 113, pp. 123-S137, 2009. 4.J. Delegido,L. Alonso, and L.G. Gonzalez, “Estimating chlorophyll content of crops from hyperspectral data using a normalized area over reflectance curve (NAOC)”. International Journal of Applied Earth Observation and Geoinformation, vol. 13, pp. 165-174, 2010. 5. P.O. Gislason, J.A. Benediktsson, and J.R. Sveinsson, “Random Forests for land cover classification”, Pattern Recognition Letters, vol. 27, pp. 294-300, 2006. 23 1. Jie Lv, Zhenguo Yan “Estimating Leaf Chlorophyll Concentration in Soybean Using Random Forests and Field Imaging Spectroscopy”IEEE Transactions On Geoscience And Remote Sensing.
- 24. 24
- 25. THANK YOU 25