Lihua Zheng

Soil nitrogen content forecasting based on real-time NIR spectroscopy

The NIR spectral absorption characteristics were analyzed for the soil samples in different layers including topsoil, subsoil and bottom soil.RSNR (relative signal-to-noise) was adopted to evaluate the filtering effect of wavelet decomposition at 1st-7th levels.After the continuum removal processing and analyzing, the sensitive wavebands (1375nm, 1520nm, 1861nm, 2100nm, 2286nm and 2387nm) were determined to predict soil TN content.Six sensitive wavebands were used to establish the regression models of soil TN content. The results showed that these sensitive wavebands could be used to predict soil TN content using real-time NIR spectra of soil. Fast and precisely estimating total nitrogen (TN) content in soil helps to promote carrying out prescription fertilization. And soil moisture is a severe interference factor in forecasting soil nitrogen content based on real-time NIR spectroscopy. This paper aims at predicting soil nitrogen content based on real-time soil spectrum through exploring pretreatment method without artificial drying and sieving soil samples. Firstly, the real-time near infrared absorbance spectra of soil samples were measured and their characteristics were analyzed. Then 1st-7th level wavelet decompositions were carried out for each soil samples real-time spectrum. RSNR (Relative Signal-to-Noise Ratio) was constructed to evaluate wavelet filtering quality at different levels, and the results indicated that low-frequency signals obtained after the 3rd level wavelet decomposing had the best performance. And then 5 soil sample groups (each group had the same moisture content but different nitrogen contents) were selected and continuum-removal method was used for processing their filtering signals. And by using the methods combined wavelet analysis and continuum removal technology, six sensitive wavebands were determined for predicting the TN content in soil, which were 1375nm, 1520nm, 1861nm, 2100nm, 2286nm and 2387nm. Finally the real-time TN content detecting models were calibrated and validated based on PLSR (Partial Least Squares Regression) and SVM (Support Vectors Machine) respectively. For the PLSR model, its calibration R2 was 0.602 and its RMSEC was 0.051mg/Kg; the validation R2 was 0.634, the RMSEP was 0.056mg/Kg and its RPD=1.838. For the SVM model, its calibration R2 reached to 0.823, the RMSEC was 0.034mg/Kg, the validation R2 reached to 0.810, the RMSEP was 0.053mg/Kg and its RPD was 2.129. It showed that, by using the proposed approach in this paper, the interference of soil moisture was mostly removed from soil real-time spectrum in the process of soil total nitrogen prediction, and the TN content regression models established by using the six sensitive wavebands had great performances in predicting soil TN content in real time.

Eliminating the interference of soil moisture and particle size on predicting soil total nitrogen content using a NIRS-based portable detector

A NIRS-based portable detector of soil TN content was developed.An algorithm was proposed to eliminate the interference of soil moisture on soil TN.A calibration was proposed to eliminate the interference of soil particle size.Combination of the two methods could well remove both the interference. Applying near infrared reflectance spectroscopy (NIRS) on farmlands can effectively estimate the total nitrogen (TN) content of soil online. We developed a NIRS-based portable detector of soil TN content that measures spectral data at 940, 1050, 1100, 1200, 1300, 1450, and 1550nm. The soil spectral data are sensitive to external environmental conditions, particularly soil moisture content and particle size. The interference of these factors on predicting soil TN content must be eliminated when using the portable detector. First, soil samples were collected from a farm in Beijing, China, and scanned using the detector to obtain their absorbance data under varying soil moisture and particle size. Second, absorbance correction method and mixed calibration set method were proposed to correct the original spectral data and to eliminate the interference of soil moisture and particle size, respectively. The absorbance of the soil sample at 1450nm exhibited a high correlation with soil moisture content. Thus, a moisture absorbance correction method (PMAI) was proposed to normalize the original spectral data into the standard spectral data and consequently eliminate the interference of soil moisture. A NIRS-based mixed calibration set based on the additivity of NIR spectra was produced with varying particle sizes, separated from the original soil samples, to eliminate the interference of soil particle size on the measurements of the portable soil TN detector. An estimation model of soil TN content was established based on the corrected absorbance data at six wavelengths (940, 1050, 1100, 1200, 1300, and 1550nm) using an algorithm of the back propagation neural network. The correlation coefficient of calibration, correlation coefficient of validation, root mean square error of calibration, root mean square error of prediction, and residual prediction deviation were used to evaluate the model. Compared with the model used the original spectral data, the accuracy and stability of the new model were significantly improved. These methods could efficiently eliminate the interference of soil moisture and particle size on predicting soil TN content.

Predicting apple sugar content based on spectral characteristics of apple tree leaf in different phenological phases

The sensitive bands of apple sugar content were found at 530-570nm and 700-720nm.Two phenological phases contribute to apple sugar accumulation higher.Apple sugar content is predictable using leaves spectra in different phenophase. Sugar degree is an important factor in determining the quality of apple. The sugar accumulation in apple fruit is closely related to fruit tree growth and development in different phases. In order to reveal the relationship between tree growth state and apple sugar content, the spectral information of apple tree leaves in different phenological phases was used to predict the fruit sugar degree. The visible and near infrared spectral reflectance of the leaves samples were measured by using a Shimadzu UV-2450 spectrograph, and the sugar content of each fruit sample growing near each leaves sample was collected and measured using laboratory methods. Then two dimensional correlation spectrum analysis was brought in, and the dynamic spectra in different phenological phases were obtained by using sugar contents as the perturbation quantity. Comprehensive observation on the spectral characteristics of leaf samples was conducted much accurately by analyzing two-dimensional correlation spectra of both synchronous and asynchronous. And then the effective spectral response bands of sugar contents and the contribution proportion to fruit sugar accumulation in different periods were investigated. And then, using the contribution proportion of each band as the single-period weighting factor, the fruit sugar sensitive wavebands were acquired. The fruit sugar content was forecasted using the sensitive bands in different phenological phases. After comparing and analyzing, it was found that the model based on parametric optimal solution of SVM showed good accuracy. The calibration R2 of the model reached to 0.8934, the RMSEC was 0.4925 Brix, the validation R2 reached to 0.8805, and its RMSEP was 0.4906 Brix. It reaches to a practical level and can be used to predict the sugar content in apple fruit.

Analysis of soil phosphorus concentration based on Raman spectroscopy

Raman spectra signature can provide structural information based on vibrational transitions of irradiated molecules. In this work, the quantity reflecting mechanism of soil phosphorus concentration was studied based on Raman spectroscopy. 15 sand soil samples with different phosphate content were made in laboratory and the Raman signatures were measured. The relationship between sand soil Phosphorus concentration and soil Raman spectra was explored. Then the effective Raman signal was extracted from the original Raman spectra by using bior4.4 wavelet packet. The relationship between sand soil phosphorus and their extracted signals was analyzed and the PLS (Partial Least Squares) model for predicting phosphorus concentration in the soil was established and compared. The maximum accuracy model comes from the extracted effective Raman spectra after the first level decomposing. The calibration R2 was close to 1 and the validation R2 reached to 0.937. It showed high potential in soil phosphorus detecting by using Raman spectroscopy.

Spectral feature extraction and modeling of soil total nitrogen content based on NIR technology and wavelet packet analysis

It is a non-destructive and real-time method to detect the soil nutrient content by using spectroscopy analysis technology. In order to isolate the effective spectral for TN content from the soil spectra effectively, the NIR model predicting TN was developed based on wavelet packet analysis. 100 soil samples were collected for calibration and validation from the field. First, using the high-precision NIR detecting instrument to scan the target and obtaining the continuous spectra of soil samples in the laboratory. Secondly, with three different orthogonal wavelets (bior4.4, db4, sym4) as the generating functions, the original signal of each soil sample was decomposed and reconstructed based on the respective wavelet packet. Then the multiple linear regression (MLR) models for TN were established based on each drawn characteristic spectrum. Finally, three models were compared and analyzed, and the model with the highest forecasting accuracy was obtained based on db4, which determined R2 reached 0.904. The research concluded that wavelet packet analysis could eliminate or substantially reduce the factors outside the parameters to the spectrum directly or indirectly, and the obstacles in establishing linear models for soil parameters were removed. It is feasible and potential to the real-time prediction of TN content.

A field information collecting system based on a wireless sensor network

Collecting environmental information of farmland is very important to the development of modern agricultural techniques. Precision farming and modern agricultural production require timely and accurate environmental information. The main task of this study was to develop a field-based Pocket Personal Computer (PC) data acquisition and processing terminal and a wireless sensor network (WSN) monitoring module, which are used to manage the working status of the field wireless sensor and to collect and record the output of the wireless sensor automatically. The ZigBee module was mainly used for wireless communication between Personal Digital Assistant (PDA) and mobile sensor nodes. The GPS module was used to obtain position information. The General Packet Radio Service (GPRS) module was used to upload the binding information of the node number, position information and agricultural information through the TCP/IP to the remote PC. The test of the whole system indicated that the system can work outside in the field. The experiment of the link quality indication (LQI) at different heights indicated that the signal would be better if the height of the ZigBee antenna was higher than 60 cm, meanwhile the LQI values would be greater than 40 at different distances. It showed that 60 cm was the least height for a better link quality.

Monitor of apple tree growth status based on spectral technology

Abstract. Spectral technology is one of the nondestructive approaches to estimate plant nitrogen status. The objective of this study is to monitor the apple tree growth status based on spectra technology in the orchards. The experiments were carried out in a Red Fuji apple orchard in Beijing. Flower/leaf samples from 15 year-on trees and 5 year-off trees were collected. The average parameters of the samples gathered from different parts of an apple tree were calculated to find the distributions of the nitrogen in the tree. The reflectance spectra of flowers/leaves from three parts (base, middle, top) of each main branch were measured. The growth environmental parameters such as the environmental temperature and humidity, the light intensity, and the leaf areas were also recorded. These could assistant us in monitor the apple tree growth status. The results showed that the correlation between the first deviation for reflectance of apple flowers and nitrogen content reached the maximum at wavelength 1201 nm. It was higher than that between the original reflectance and nitrogen content. The reflectance spectra of apple trees changed significantly at different stages. Leaves from the top of the branch had higher reflectance than the other parts of the branch. The adjusted R 2 of the linear regression models for predicting leaf nitrogen content by using the combination of reflectance at 712 nm and 800 nm could reach to 0.482 while the RMSE was 0.238 (n=55).

Development of a smart greenhouse monitoring system based on WSN, 3G and PLC technologies

Abstract. It is very convenient for farmers to control the devices located in greenhouses remotely and precisely. A smart greenhouse monitoring system was developed based on WSN, 3G and PLC technologies. It consists of a WSN, a 3G communication unit, a PLC control unit and devices (air fans and roller shutter door) in each greenhouse and a monitoring application running on the greenhouse server. The system manages greenhouse for each registered farmer. The WSN is developed using ZigBee technology and each WSN node integrates environment humidity sensor and temperature sensor, and the power for each sensor can be disenabled by the wireless node itself after sending data for 10 times and enabled when receiving the data collecting command from the coordinator. There is a ZigBee coordinator in each greenhouse and it is integrated with a 3G module and a PLC unit. Furthermore, the MCU in ZigBee coordinator is used as the CU for the integrated setup. The devices installed in each greenhouse are connected with the PLC unit and can be triggered and controlled by the PLC module. The WSN collects all of data in given interval. The 3G module is responsible for establishing the connection with the greenhouse server, uploading the collected data to the server and downloading the commands from the server. The PLC unit receives the commands forwarded by the CU and controls each device to execute the specific action.

Prediction of Water Chlorophyll-a Content Based on Multi-scale Spectral Analysis

Abstract. In this research, samples were collected from JiangSu Province of China. The visible and near infrared spectral reflectance of the water samples were measured. At the same time, the Chlorophyll-a content of water for each sample was measured in the laboratory. Then the spectral characteristics of the water samples were analyzed and the results showed that with chlorophyll-a concentration increasing, spectral transmittance decreased gradually. There was an apparent transmission valley at 676nm. Then multiscale spectral analysis aimed at predicting the chlorophyll-a content in water were conducted, which were consisted of all-wavebands prediction and specific sensitive wavebands prediction. For all-wavebands spectral analysis, this research adopted the SVM technology to establish the regression model based on the all-bands transmittance between 331~900nm. For the specific sensitive wavebands prediction, spectral correlation coefficient analysis was carried out to analyze the sensitive transmittance band of Chlorophyll-a in water. Comprehensive observation on the correlation coefficient curve, the significant correlation could be seen at around 365nm, 550nm and 856nm. It implies that water chlorophyll-a concentration linear forecasting model can be established by the sensitive wavebands selected above. After comparing, the results indicated that, 1) the accuracy of all-wavebands regression model reached to a quite high level, because there is no losing information from the spectra. 2) the sensitive wavebands regression model could also predict the chlorophyll-a content in water practically, and the model was more suitable for embedding in the detector and applying to the agricultural production.

Development of crop monitoring system using 2-channel CCD image sensor

Abstract. A feasible multi-spectral acquisition system was developed for crop monitoring. The system included multi-spectral image acquisition hardware and controlling software. The multi-spectral image acquisition device was designed with a multi-spectral 2-channel CCD sensor to measure multi-spectral images in visible (red (R), green (G) and blue (B) ) and near infrared (NIR) waveband. A panel industrial control computer was applied as the system platform. The QuadMVCL2GE was used as a converter to connect the CCD sensor and controlling computer. The output of 2-CCD sensor with Camera Link was converted into Gigabit Ethernet communication protocol. The improvement of the developed system was considered to simplify the communication protocol conversion. Thus, a new simple multi-spectral image acquisition system was developed. And the controlling software interface was refined coordinately. When the system was connected, it could work following image acquisition, image display and storage, and image processing. The multi-spectral image of tomato plant canopy was collected and processed following image enhancement, canopy segmentation and vegetation index calculation. The average gray value (R, G, B and NIR) and the vegetation indices (RVI, NDVI, et al.) widely used in remote sensing were selected as the parameters for tomato chlorophyll content monitoring. The results showed that the improved multi-spectral image acquisition system could be used to evaluate the growth of tomato in the greenhouse. It provides an efficient tool to obtain crop information for precision agriculture.