A near-infrared (NIR) hyperspectral imaging system was developed in this study. NIR hyperspectral imaging combined with multivariate data analysis was applied to identify rice seed cultivars. Spectral data was exacted from hyperspectral images. Along with Partial Least Squares Discriminant Analysis (PLS-DA), Soft Independent Modeling of Class Analogy (SIMCA), K-Nearest Neighbor Algorithm (KNN) and Support Vector Machine (SVM), a novel machine learning algorithm called Random Forest (RF) was applied in this study. Spectra from 1,039 nm to 1,612 nm were used as full spectra to build classification models. PLS-DA and KNN models obtained over 80% classification accuracy, and SIMCA, SVM and RF models obtained 100% classification accuracy in both the calibration and prediction set. Twelve optimal wavelengths were selected by weighted regression coefficients of the PLS-DA model. Based on optimal wavelengths, PLS-DA, KNN, SVM and RF models were built. All optimal wavelengths-based models (except PLS-DA) produced classification rates over 80%. The performances of full spectra-based models were better than optimal wavelengths-based models. The overall results indicated that hyperspectral imaging could be used for rice seed cultivar identification, and RF is an effective classification technique.

Rice grain moisture has a great impact on th production and storage storage quality of rice. The main objective of this study was to design and develop a rapid-detection sensor for rice grain moisture based on the Near-infrared spectroscopy (NIR) characteristic band, aiming to realize its accurate and on-line measurement. In this paper, the NIR spectral information of grain samples with different moisture content was obtained using a portable NIR spectrometer. Then, the partial least squares (PLS) and competitive adaptive reweighted squares (CARS) were applied to model and analyze the spectral data to find the rice grain moisture NIR spectroscopy. As a result, the 1450 nm band was sensitive to the rice grain moisture and a rapid-detection sensor was developed with a 1450 nm light emitting diode (LED) light source, InGaAs photodiode, lens and filter, whose basic principle is to establish the relationship between the rice grain moisture and the measured voltage signal. To evaluate the sensor performance, rice grain samples with 13–30% moisture content were detected, the coefficient of determination R² was 0.936, and the sum of squares for error (SSE) was 23.44. It is concluded that this study provides a spectroscopic measuring method, as well as developing an effective and accurate sensor for the rapid determination of rice grain moisture, which is of great significance for monitoring the quality of rice grain during its production, transportation and storage process.

Soil nitrogen content is one of the important growth nutrient parameters of crops. It is a prerequisite for scientific fertilization to accurately grasp soil nutrient information in precision agriculture. The information about nutrients such as nitrogen in the soil can be obtained quickly by using a near-infrared sensor. The data can be analyzed in the detection process, which is nondestructive and non-polluting. In order to investigate the effect of soil pretreatment on nitrogen content by near infrared sensor, 16 nitrogen concentrations were mixed with soil and the soil samples were divided into three groups with different pretreatment. The first group of soil samples with strict pretreatment were dried, ground, sieved and pressed. The second group of soil samples were dried and ground. The third group of soil samples were simply dried. Three linear different modeling methods are used to analyze the spectrum, including partial least squares (PLS), uninformative variable elimination (UVE), competitive adaptive reweighted algorithm (CARS). The model of nonlinear partial least squares which supports vector machine (LS-SVM) is also used to analyze the soil reflectance spectrum. The results show that the soil samples with strict pretreatment have the best accuracy in predicting nitrogen content by near-infrared sensor, and the pretreatment method is suitable for practical application.

The terahertz (THz) spectra of rapeseed leaves with different water content (WC) were investigated. The transmission and absorption spectra in the range of 0.3–2 THz were measured by using THz time-domain spectroscopy. The mean transmittance and absorption coefficients were applied to analyze the change regulation of WC. In addition, the Savitzky-Golay method was performed to preprocess the spectra. Then, the partial least squares (PLS), kernel PLS (KPLS), and Boosting-PLS were conducted to establish models for predicting WC based on the processed transmission and absorption spectra. Reliable results were obtained by these three methods. KPLS generated the best prediction accuracy of WC. The prediction coefficient correlation (Rval) and root mean square error (RMSEP) of KPLS based on transmission were Rval = 0.8508, RMSEP = 0.1015, and that based on absorption were Rval = 0.8574, RMSEP = 0.1009. Results demonstrated that THz spectroscopy combined with modeling methods provided an efficient and feasible technique for detecting plant physiological information.

Thiabendazole is widely used in sclerotium blight, downy mildew and black rot prevention and treatment in rape. Accurate monitoring of thiabendazole pesticides in plants will prevent potential adverse effects to the Environment and human health. Surface Enhanced Raman Spectroscopy (SERS) is a highly sensitive fingerprint with the advantages of simple operation, convenient portability and high detection efficiency. In this paper, a rapid determination method of thiabendazole pesticides in rape was conducted combining SERS with chemometric methods. The original SERS were pretreated and the partial least squares (PLS) was applied to establish the prediction model between SERS and thiabendazole pesticides in rape. As a result, the SERS enhancing effect based on silver Nano-substrate was better than that of gold Nano-substrate, where the detection limit of thiabendazole pesticides in rape could reach 0.1 mg/L. Moreover, 782, 1007 and 1576 cm⁻¹ could be determined as thiabendazole pesticides Raman characteristic peaks in rape. The prediction effect of thiabendazole pesticides in rape was the best ( $R_p^2$ = 0.94, RMSEP = 3.17 mg/L) after the original spectra preprocessed with 1st-Derivative, and the linear relevance between thiabendazole pesticides concentration and Raman peak intensity at 782 cm⁻¹ was the highest (R² = 0.91). Furthermore, five rape samples with unknown thiabendazole pesticides concentration were used to verify the accuracy and reliability of this method. It was showed that prediction relative standard deviation was 0.70–9.85%, recovery rate was 94.71–118.92% and t value was −1.489. In conclusion, the thiabendazole pesticides in rape could be rapidly and accurately detected by SERS, which was beneficial to provide a rapid, accurate and reliable scheme for the detection of pesticides residues in agriculture products.