Hyperspectral image (HSI) contains rich spatial and spectral information, which has been widely used in resource exploration, ecological environment monitoring, land cover classification and target recognition. However, the nonlinearity of HSI data and the strong correlation between bands also bring difficulties and challenges to HSI application. In particular, the limited available hyperspectral training samples will lead to the classification accuracy cannot be improved. Therefore, making full use of the advantages of HSI data, through algorithms and strategies to solve the limited training samples, high-dimensional HSI data and effective classification method, so as to improve the classification accuracy. This paper reviews the research results of the feature extraction methods and classification methods of HSI classification in recent years. In addition, this paper expounds five kinds of small sample strategies, and solves the problem of small sample in HSI classification from different angles. Small sample strategy will be the focus of HSI classification research in the future. To solve the problem of small sample classification can greatly promote the application of HSI.

Changes in land cover will cause the changes in the climate and environmental characteristics, which has an important influence on the social economy and ecosystem. The main form of land cover is different types of soil. Compared with traditional methods, visible and near-infrared spectroscopy technology can classify different types of soil rapidly, effectively, and nondestructively. Based on the visible near-infrared spectroscopy technology, this paper takes the soil of six different land cover types in Qingdao, China orchards, woodlands, tea plantations, farmlands, bare lands, and grasslands as examples and establishes a convolutional neural network classification model. The classification results of different number of training samples are analyzed and compared with the support vector machine algorithm. Under the condition that Kennard–Stone algorithm divides the calibration set, the classification results of six different soil types and single six soil types by convolutional neural network are better than those by the support vector machine. Under the condition of randomly dividing the calibration set according to the proportion of 1/3 and 1/4, the classification results by convolutional neural network are also better. The aim of this study is to analyze the feasibility of land cover classification with small samples by convolutional neural network and, according to the deep learning algorithm, to explore new methods for rapid, nondestructive, and accurate classification of the land cover.

Although visible-near-infrared spectroscopy can rapidly and accurately determine soil nutrients without sample destruction, some problems remain unresolved, such as the mismatch of the established spectral model with different types of samples, limiting the wide application of this technology. Here, we took riverside and mountain soils as examples to explore the calibration transfer between two different types of soils by the WMPDS-S/B algorithm (wavelet multiscale piecewise direct standardization combined with Slope/Bias correction method) and by adding new samples. The predicted TN and TC concentrations improved significantly after being transformed. Compared with adding new samples, the WMPDS-S/B algorithm obtained more accurate results. The average relative errors dropped from 440.2% (without transformation) to approximately 6% for TN and from 342.0% to approximately 7% for TC. The maximum relative errors were reduced from 538.1% to less than 20% for TN and from 403.7% to less than 20% for TC. The RMSEP decreased from 2.42 to approximately 0.04 for TN and from 15.74 to approximately 0.4 for TC. The WMPDS-S/B algorithm had advantages in selecting fewer known samples and obtaining better prediction results. In contrast to past studies, which resolved the calibration transfer between different spectrometers and the measurement environment for the same samples, our study resolved the calibration transfer between different types of samples under the same spectrometer and the measurement environment. The former could only be used for correction among instruments, while the latter fundamentally solved the problem of model sharing across different samples.

The calibration transfer between instruments is mainly aimed at the calibration transfer between normal spectrometers. There are few studies on the calibration transfer of soil nutrient concentration from a hyperspectral camera to a normal spectrometer. In this paper, 164 soil samples from three regions in Qingdao, China, were collected. The spectral data of normal spectrometer and hyperspectral camera and the concentration of total carbon and nitrogen were obtained. And then, the models of soil total carbon and nitrogen content were established by using the spectral data of a normal spectrometer. The hyperspectral data were transferred by a variety of methods, such as single conventional calibration transfer algorithm, combination of multiple calibration transfer algorithms, and calibration transfer algorithm after spectral pretreatment. The transferred hyperspectral data were predicted by the total carbon and total nitrogen concentration model established by using a normal spectrometer. The absolute coefficients Rt2 and root mean square error of prediction (RMSEP) were used to evaluate the prediction performance after calibration transfer. After trying many calibration transfer methods, the prediction performance of calibration transfer by the Repfile-PDS and Repfile-SNV methods was the best. In the calibration transfer of the Repfile-PDS method, when the number of PDS windows was 27 and the number of standard data was 40, the Rt2 and the RMSEP of TC concentration were 0.627 and 2.351. When the number of PDS windows was 25 and the number of standard data was 100, the Rt2 and the RMSEP of TN concentration were 0.666 and 0.297. In the calibration transfer of the Repfile-SNV method, when the number of TC and TN standard data was 120, the Rt2 was the largest, 0.701 and 0.722, respectively, and the RMSEP was 2.880 and 0.399, respectively. After the hyperspectral data were calibration transferred by the above algorithms, they could be predicted by the soil TC and TN concentration model established by using a normal spectrometer, and better prediction results can be obtained. The solution of the calibration transfer of soil nutrient concentration from the hyperspectral camera to the normal spectrometer provides a powerful basis for rapid prediction of a large number of image information data collected by using a hyperspectral camera. It greatly reduces the workload and promotes the application of hyperspectral camera in quantitative analysis and rapid measurement technology.