Nitrogen nitrates play a significant role in the soil\'s nutrient cycle, and near-infrared spectroscopy can efficiently and accurately detect the content of nitrate-nitrogen in the soil. Accordingly, it can provide a scientific basis for soil improvement and agricultural productivity by deeply examining the cycle and transformation pattern of nutrients in the soil. To investigate the impact of drying temperature on NIR soil nitrogen detection, soil samples with different N concentrations were dried at temperatures of 50 °C, 65 °C, 80 °C, and 95 °C, respectively. Additionally, soil samples naturally air-dried at room temperature (25 °C) were used as a control group. Different drying times were modified based on the drying temperature to completely eliminate the impact of moisture. Following data collection with an NIR spectrometer, the best preprocessing method was chosen to handle the raw data. Based on the feature bands chosen by the RFFS, CARS, and SPA methods, two linear models, PLSR and SVM, and a nonlinear ANN model were then established for analysis and comparison. It was found that the drying temperature had a great effect on the detection of soil nitrogen by near-infrared spectroscopy. In the meantime, the SPA-ANN model simultaneously yielded the best and most stable accuracy, with Rc2 = 0.998, Rp2 = 0.989, RMSEC = 0.178 g/kg, and RMSEP = 0.257 g/kg. The results showed that NIR spectroscopy had the least effect and the highest accuracy in detecting nitrogen at 80 °C soil drying temperature. This work provides a theoretical foundation for agricultural production in the future.

Plastic debris are ubiquitous in soil and bring severe threatening to environment and ecosystem. It is of great significance to extensively investigate the plastic pollution level in soil. An ultra-portable Near-infrared (NIR) sensor was used to detect plastic pollution level in soil. Soil samples were collected from three different regions and artificially polluted in two degrees (10-1.5% and 0.7-0.15%). Here, instead of constructing detection models for specific soil region, transfer learning approaches were explored to build classification model which could evaluate plastic pollution level in different soil regions simultaneously. The transfer learning algorithms, Manifold Embedded Distribution Alignment (MEDA) and Transfer Component Analysis (TCA), were employed for transfer learning model construction. Supporting Vector Machine (SVM) models were calibrated for transferability analysis and comparison. MEDA transferable models achieved the average classification accuracy of 97.78% in source soil regions and 79.52% in target soil regions. The average accuracy of TCA based models and conventional SVM models were equivalent to each other and lower than MEDA models. Besides, the average running time of MEDA method (0.70 s) was much lower than TCA based method (21.90 s) and conventional SVM models (41.38 s). Overall, the results indicated that transfer learning approaches especially MEDA method could work in a more efficient manner than that of conventional multivariate analysis. The ultra-portable NIR sensor in combination with MEDA transfer learning algorithm as modelling method was a promising solution for low-cost and efficient field detection of plastic contaminated level in soil.

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.