Nitrogen (N) is important for the growth of crops. Estimating leaf nitrogen concentration (LNC) accurately and nondestructively is important for precision agriculture, reduces environmental pollution, and helps model global carbon and N cycles. Leaf reflectance, especially in the visible and near-infrared regions, has been identified as a useful indicator of LNC. Except reflectance passively acquired by spectrometers, the newly developed multispectral LiDAR and hyperspectral LiDAR provide possibilities for measuring leaf spectra actively. The regression relationship between leaf reflectance spectra and rice (Oryza sativa) LNC relies greatly on the algorithm adopted. It would be preferable to find one algorithm that performs well with respect to passive and active leaf spectra. Thus, this study assesses the influence of six popular linear and nonlinear methods on rice LNC retrieval, namely, partial least-square regression, least squares boosting, bagging, random forest, back-propagation neural network (BPNN), and support vector regression of different types/kernels/parameter values. The R², root mean square error and relative error in rice LNC estimation using these different methods were compared through the passive and active spectral measurements of rice leaves of different varieties at different locations and time (Yongyou 4949, Suizhou, 2014, Yangliangyou 6, Wuhan, 2015). Results demonstrate that BPNN provided generally satisfactory performance in estimating rice LNC using the three kinds of passive and active reflectance spectra.

Target classification techniques using spectral imagery and light detection and ranging (LiDAR) are widely used in many disciplines. However, none of the existing methods can directly capture spectral and 3D spatial information simultaneously. Multispectral LiDAR was proposed to solve this problem as its data combines spectral and 3D spatial information. Point-based classification experiments have been conducted with the use of multispectral LiDAR; however, the low signal to noise ratio creates salt and pepper noise in the spectral-only classification, thus lowering overall classification accuracy. In our study, a two-step classification approach is proposed to eliminate this noise during target classification: routine classification based on spectral information using spectral reflectance or a vegetation index, followed by neighborhood spatial reclassification. In an experiment, a point cloud was first classified with a routine classifier using spectral information and then reclassified with the k-nearest neighbors (k-NN) algorithm using neighborhood spatial information. Next, a vegetation index (VI) was introduced for the classification of healthy and withered leaves. Experimental results show that our proposed two-step classification method is feasible if the first spectral classification accuracy is reasonable. After the reclassification based on the k-NN algorithm was combined with neighborhood spatial information, accuracies increased by 1.50–11.06%. Regarding identification of withered leaves, VI performed much better than raw spectral reflectance, with producer accuracy increasing from 23.272% to 70.507%.

The abilities of multispectral LiDAR (MSL) as a new high-potential active instrument for remote sensing have not been fully revealed. This study demonstrates the potential of using the spectral and spatial features derived from a novel MSL to discriminate surface objects. Data acquired with the MSL include distance information and the intensities of four wavelengths at 556, 670, 700, and 780 nm channels. A support vector machine was used to classify diverse objects in the experimental scene into seven types: wall, ceramic pots, Cactaceae, carton, plastic foam block, and healthy and dead leaves of E. aureum. Different features were used during classification to compare the performance of different detection systems. The spectral backscattered reflectance of one wavelength and distance represented the features from an equivalent single-wavelength LiDAR system; reflectance of the four wavelengths represented the features from an equivalent multispectral image with four bands. Results showed that the overall accuracy of using MSL data was as high as 88.7%, this value was 9.8%–39.2% higher than those obtained using a single-wavelength LiDAR, and 4.2% higher than for multispectral image.

Multispectral lidar has become a promising technology with the rise in capability of 3-D spectral imaging. However, the precise acquisition of spectral information is interfered by measurement geometry, namely, incidence angle and detection distance. These issues may cause discrepancy within the spectral information, thus limiting the classification capabilities of multispectral lidar. To fill this gap, a hue-saturation-intensity (HSI) color space-based method for multispectral lidar classification is proposed in this study. The proposed scheme does not require radiometric calibration, as the HSI color space is robust to spectral intensity variations within a single target. In this method, spectral data are transformed from red-green-blue (RGB) color space to HSI color space. The three components of the HSI color space are inputted for the classification. Then, a reference target-based radiometric calibration is conducted for comparison. The complex indoor scene and the random forest classifier are used for the validation. The classification results of using raw RGB data, raw HSI data, calibrated RGB data, and calibrated HSI data are compared. Results show that the raw HSI data outperform the raw RGB data in terms of classification accuracy. In particular, the raw HSI data can correct the classification error caused by the measurement geometry more effectively than the calibrated RGB data. The improvement resulting from using the HSI color space is demonstrated by both the three-wavelength multispectral lidar and the 32-channel multispectral lidar. That indicates that HSI color space is a promising tool for enhancing the classification capability of multispectral lidar.

Biochemicals, such as chlorophyll (Chl) and nitrogen (N), are closely related to photosynthesis process of vegetation. Their accurate estimation is an important topic in remote sensing of vegetation. Previous studies mainly focused on Chl-N content inversion in leaf and canopy level, and few cared about their 3-D distributions, which was also an important indicator for the growth status of vegetation (GSV). Hyperspectral LiDAR (HSL) is a novel active remote sensing technology, which has target-sensitive band with hyperspectra resolution. Its 3-D point cloud data simultaneously contains rich spectral and precise geometrical characteristics of the target. This work aims to apply HSL data on 3-D Chl-N content mapping in vegetation through constructing HSL-based spectral indices (SIs). Except for following the SI forms of previous works, the normalized differential vegetation index and ratio index (RI) with four broadbands in an HSL spectral space were successively proposed to invert Chl-N content for the whole vegetation based on the artificial neural network (ANN) method. These four broadbands were transformed based on the relative spectral response curve of detector and the feature weights (FWs) of multiwavelength, respectively. Results show that most HSL-based ANN models can accurately invert Chl-N content with a mean R² of >0.75, and some that fusing broadband data with convolution transformation, namely the FW-based RI, can even obtain a model R² of 0.84 for N content inversion. Thus, HSL can be efficiently applied to 3-D Chl-N content mapping of vegetation and has great potential in GSV monitoring.

Leaf nitrogen concentration (LNC) is a significant indicator of crops growth status, which is related to crop yield and photosynthetic efficiency. Laser-induced fluorescence is a promising technology for LNC estimation and has been widely used in remote sensing. The accuracy of LNC monitoring relies greatly on the selection of fluorescence characteristics and the number of fluorescence characteristics. It would be useful to analyze the performance of fluorescence intensity and ratio characteristics at different wavelengths for LNC estimation. In this study, the fluorescence spectra of paddy rice excited by different excitation light wavelengths (355 nm, 460 nm, and 556 nm) were acquired. The performance of the fluorescence intensity and fluorescence ratio of each band were analyzed in detail based on back-propagation neural network (BPNN) for LNC estimation. At 355 nm and 460 nm excitation wavelengths, the fluorescence characteristics related to LNC were mainly located in the far-red region, and at 556 nm excitation wavelength, the red region being an optimal band. Additionally, the effect of the number of fluorescence characteristics on the accuracy of LNC estimation was analyzed by using principal component analysis combined with BPNN. Results demonstrate that at least two fluorescence spectral features should be selected in the red and far-red regions to estimate LNC and efficiently improve the accuracy of LNC estimation.