Search

Hyperspectral imaging is a promising technique for nondestructive sensing of multiple quality attributes of apple fruit. This research evaluated and compared different mathematical models to extract effective wavelengths for measurement of apple soluble solids content (SSC) based on near infrared (NIR) hyperspectral imaging over the spectral region of 1000–2500 nm. A total of 160 samples were prepared for the calibration (n = 120) and prediction (n = 40) sets. Competitive adaptive reweighted sampling (CARS), successive projections algorithm (SPA), random frog (RF), and CARS-SPA, CARS-RF combined algorithms were used for extracting effective wavelengths from hyperspectral images of apples, respectively. Based on the selected effective wavelengths, different models were built and compared for predicting SSC of apple using partial least squares (PLS) and least squared support vector regression (LS-SVR). Among all the models, the models based on the ten effective wavelengths selected by CARS-SPA achieved the best results, with Rp, RMSEP of 0.907, 0.479 °Brix for PLS and 0.917, 0.453 °Brix for LS-SVR, respectively. The overall results indicated that CARS-SPA can be used for selecting the effective wavelengths from hyperspectral data. Both PLS and LS-SVR can be applied to develop calibration models to predict apple SSC. Furthermore, the wavelengths selected by CARS-SPA algorithm has a great potential for online detection of apple SSC.

The effect of variation of fruit orientation on online prediction of soluble solids content (SSC) of ‘Fuji’ apples based on visible and near-infrared (Vis/NIR) spectroscopy was studied. The diffuse reflectance spectra in 550–950 nm were collected with a designed online system in six orientations: stem-calyx axis vertical with stem upward (T1) and stem downward (T5), 45° between stem-calyx axis and horizontal with stem slope upward (T2) and stem slope downward (T4), stem-calyx axis horizontal with stem towards computer side lights (T3), stem-calyx axis horizontal with stem towards belt movement direction (T6). The 180 samples with SSC range of 8.00–13.60°Brix were divided into 135 of calibration set with 1.09 standard deviation (S.D.) and 45 of prediction set with 0.85 S.D. The signal-to-noise ratio (SNR) and area change rate (ACR) were used to evaluate the stability of collected spectra. After the comparison of different preprocessing methods, partial least squares (PLS) and least squares-support vector machine (LS-SVM) were used to develop compensation models of SSC for each orientation separately (local models) and all orientations (global model), respectively. Finally, competitive adaptive reweighted sampling (CARS), successive projection algorithm (SPA), and their combination were used to select the effective wavelengths (EWs), respectively. Results showed that T1 performed better for our system and influence of measurement orientation on spectra greatly affected SSC prediction accuracy. Comparatively, global model was insensitive to fruit orientation variation. 37 EWs selected by CARS-SPA-PLS model after Savitzky-Golay smoothing in all orientations achieved better results with rp and RMSEP of 0.815, 0.818, 0.837, 0.731, 0.807, 0.842 and 0.487, 0.484, 0.460, 0.573, 0.497, 0.453°Brix, respectively. Generally, global model with EWs could be promisingly used for online SSC prediction of apple.

Moisture content (MC) is one of the most important factors for assessment of seed quality. However, the accurate detection of MC in single seed is very difficult. In this study, single maize seed was used as research object. A long-wave near infrared (LWNIR) hyperspectral imaging system was developed for acquiring reflectance images of the embryo and endosperm side of maize seed in the spectral range of 930–2548 nm, and the mixed spectra were extracted from both side of maize seeds. Then, Full-spectrum models were established and compared based on different types of spectra. It showed that models established based on spectra of the embryo side and mixed spectra obtained better performance than the endosperm side. Next, a combination of competitive adaptive reweighted sampling (CARS) and successive projections algorithm (SPA) was proposed to select the most effective wavelengths from full-spectrum data. In order to explore the stableness of wavelength selection algorithm, these methods were used for 200 independent experiments based on embryo side and mixed spectra, respectively. Each selection result was used as input of partial least squares regression (PLSR) and least squares support vector machine (LS-SVM) to build calibration models for determining the MC of single maize seed. Results indicated that the CARS-SPA-LS-SVM model established with mixed spectra was optimal for MC prediction in all models by considering the accuracy, stableness and complexity of models. The prediction accuracy of CARS-SPA-LS-SVM model is Rpre = 0.9311 ± 0.0094 and RMSEP = 1.2131 ± 0.0702 in 200 independent assessment. The overall study revealed that the long-wave near infrared hyperspectral imaging can be used to non-invasively and fast measure the MC in single maize seed and a robust and accurate model could be established based on CARS-SPA-LS-SVM method coupled with mixed spectral. These results can provide a useful reference for assessment of other internal quality attributes (such as starch content) of single maize seed.

As an essential factor in quality assessment of maize seeds, variety purity profoundly impacts final yield and farmers' economic benefits. In this study, a novel method based on Raman hyperspectral imaging system was applied to achieve variety classification of coated maize seeds. A total of 760 maize seeds including 4 different varieties were evaluated. Raman spectral data of 400-1800 cm

The rapid and non-destructive detection of moisture in withering leaves is an unsolved problem because the leaves are stacked together and have random orientation. To address this issue, this study aimed to establish more robust and accurate models. The performance of front side, back side and multi-region models were compared, and the front side model showed the worst transferability. Therefore, five effective wavelength (EW) selection algorithms were combined with a successive projection algorithm (SPA) to select EWs. It was found that the shuffled frog leaping algorithm (SFLA) combined with SPA was the best method for the front side model for moisture analyses. Based on the selected EWs, the extreme learning machine (ELM) became the model with the best self-verification result. Subsequently, moisture distribution maps of withering leaves were successfully generated. Considering the processing demand of withering leaves, local region models developed based on partial least squares and the SFLA-SPA method were applied to predict the moisture of withering leaves in the local and stacked region. The results showed that the RPD, R

In this study Vis/NIR spectroscopy was applied to evaluate soluble solids content (SSC) of tomato. A total of 168 tomato samples with five different maturity stages, were measured by two developed systems with the wavelength ranges of 500–930 nm and 900–1400 nm, respectively. The raw spectral data were pre-processed by first derivative and standard normal variate (SNV), respectively, and then the effective wavelengths were selected using competitive adaptive reweighted sampling (CARS) and random frog (RF). Partial least squares (PLS) and least square-support vector machines (LS-SVM) were employed to build the prediction models to evaluate SSC in tomatoes. The prediction results revealed that the best performance was obtained using the PLS model with the optimal wavelengths selected by CARS in the range of 900–1400 nm (Rp = 0.820 and RMSEP = 0.207 °Brix). Meanwhile, this best model yielded desirable results with Rp and RMSEP of 0.830 and 0.316 °Brix, respectively, in 60 samples of the independent set. The method proposed from this study can provide an effective and quick way to predict SSC in tomato.

Currently, the detection of blueberry internal bruising focuses mostly on single hyperspectral imaging (HSI) systems. Attempts to fuse different HSI systems with complementary spectral ranges are still lacking. A push broom based HSI system and a liquid crystal tunable filter (LCTF) based HSI system with different sensing ranges and detectors were investigated to jointly detect blueberry internal bruising in the lab. The mean reflectance spectrum of each berry sample was extracted from the data obtained by two HSI systems respectively. The spectral data from the two spectroscopic techniques were analyzed separately using feature selection method, partial least squares-discriminant analysis (PLS-DA), and support vector machine (SVM), and then fused with three data fusion strategies at the data level, feature level, and decision level. The three data fusion strategies achieved better classification results than using each HSI system alone. The decision level fusion integrating classification results from the two instruments with selected relevant features achieved more promising results, suggesting that the two HSI systems with complementary spectral ranges, combined with feature selection and data fusion strategies, could be used synergistically to improve blueberry internal bruising detection. This study was the first step in demonstrating the feasibility of the fusion of two HSI systems with complementary spectral ranges for detecting blueberry bruising, which could lead to a multispectral imaging system with a few selected wavelengths and an appropriate detector for bruising detection on the packing line.

Maize is an important food crop in the world and it is used in many fields. The classification of maize seed maturity is of great value because it could increase the yield. In this study, near-infrared hyperspectral imaging (NIR-HSI) was employed to explore the maturity classification of maize seeds. In order to observe the influence of spectra of different positions in maize seed for modeling, the hyperspectral images of embryo and endosperm sides of maize seeds were collected in the spectral range of 1000–2300 nm. The average spectra of the embryo side (T1) and endosperm (T2) side were extracted from hyperspectral images, and then, the average spectra of both sides of maize seed (T3) were also calculated. T1, T2 and T3 spectra were used to build calibration models for maturity classification, respectively. And T1 and T2 spectra were imported into these developed classification models, and the classification accuracy of two types of spectra in the model was used to evaluate model applicability. These modeling methods including partial least square discriminant analysis (PLS-DA), decision tree (DT) and adaptive boosting (AdaBoost) methods. The principal component analysis (PCA) was applied to select feature wavelengths, common peaks and valleys in the loading curves of PC1 and PC2 were regarded as feature wavelengths. In order to reduce the influence of division of the calibration set, 50 randomized independent trials were carried out, and the average accuracy and stableness were used to evaluate the performance of models. Comparing among all models, PLS-DA model based on feature wavelengths selected by T2 spectra obtained the optimal results. When T1 and T2 were used as input to the optimal model, the classification accuracy was 98.7% and 100%, respectively. These results demonstrate the potential of the hyperspectral imaging technology for the rapid and accurate classification of maize seed maturity, and the feature wavelengths selected from the endosperm side combined with PLS-DA algorithm could establish a stable model.

Development of the online and nondestructive technologies for inspecting and grading the quality of fruit in the postharvest period can improve industry competitiveness and profitability. The effect of fruit temperature, diameter and weight on online evaluation system of soluble solids content (SSC) of ‘Ya’ pears using visible/near infrared (Vis/NIR) spectroscopy was studied. To establish calibration models, partial least square (PLS) regression and least squares-support vector machine (LS-SVM) were employed in 630–900 nm and two fruit orientations (stem-calyx axis vertical with stem upward (T1), stem-calyx axis horizontal with stem towards belt moving direction (T2)), respectively. After pretreatments of Savitzky-Golay smoothing (SGS), multiplicative scattering correction (MSC), standard normal variate (SNV), and competitive adaptive reweighted sampling (CARS) for effective wavelength (EWs) selection, models were optimized and compared to evaluate calibration strategies. 36 EWs using PLS (rp = 0.89, RMSEP = 0.56) with the consideration of diameter (T1) and 34 EWs using LS-SVM (rp = 0.90, RMSEP = 0.57) with the consideration of temperature and diameter (T2) were finally selected, respectively. The fusion information of temperature and diameter showed beneficial effect and the best prediction results based on the designed online Vis/NIR half-transmittance system after MSC and 7-SGS for SSC evaluation of pears using LS-SVM, which would be effective to simplify models and promote computing efficiency and further make this proposed nondestructive detection technique have the practical application.

Two hyperspectral imaging (HSI) systems, visible/near infrared (Vis/NIR, 304–1082 nm) and short wave infrared (SWIR, 930–2548 nm), were used for the first time to comprehensively predict the changes in quality of wheat seeds based on three vigour parameters: germination percentage (GP, reflecting the number of germinated seedling), germination energy (GE, reflecting the speed and uniformity of seedling emergence), and simple vigour index (SVI, reflecting germination percentage and seedling weight). Each sample contained a small number of wheat seeds, which were obtained by high temperature and humidity-accelerated aging (0, 2, and 3 days) to simulate storage. The spectra of these samples were collected using HSI systems. After collection, each seed sample underwent a standard germination test to determine their GP, GE, and SVI. Then, several pretreatment methods and the partial least-squares regression algorithm (PLS-R) were used to establish quantitative models. The models for the Vis/NIR region obtained excellent performance, and most effective wavelengths (EWs) were selected in the Vis/NIR region by the successive projections algorithm (SPA) and regression coefficients (RC). Subsequently, PLS-R-RC models using selected wavebands (sixteen wavebands for GP, 14 wavebands for GE, and 16 wavebands for SVI) exhibited similar performance to the PLS-R models based on the full wavebands. The best R2 results obtained in the simplified models' prediction sets were 0.921, 0.907, and 0.886, with RMSE values of 4.113%, 5.137%, and 0.024, for GP, GE, and SVI, respectively. Distribution maps of GP, GE, and SVI were produced by applying these simplified PLS models. By interpreting the EWs and building prediction models, soluble protein and sugar content were demonstrated to have a relationship with spectral information. In summary, the present results lay a foundation towards the development of a significantly simpler, more comprehensive, and non-destructive hyperspectral-based sorting system for determining the vigour of wheat seeds.

A novel method for discriminating the varieties of sweet maize seeds was developed on the basis of hyperspectral imaging technology in the visible and near-infrared (Vis–NIR) region (326.7–1098.1 nm). First, the Vis–NIR hyperspectral images of nine varieties of sweet maize seeds were obtained with the orientations of germ up and down. Second, Savitzky–Golay (SG) smoothing and first derivative (FD) methods were used to highlight the differences of different maize seeds. Finally, a variety discrimination model was established by support vector machine (SVM) based on the effective wavelengths extracted by competitive adaptive reweighted sampling (CARS) algorithm. Additionally, the performance of other six comparative algorithms including successive projections algorithm (SPA), principal component analysis (PCA), factor analysis (FA), random projection (RP), independent component analysis (ICA), and t-distributed stochastic neighbor embedding (t-SNE) were compared with CARS. The classification models of SVM was also compared with Naive Bayes (NB), K-nearest neighbors (KNN), artificial neural networks (ANN), decision tree (DT), linear discriminant analysis (LDA) and logistic regression (LR) algorithms. Results showed that the SG + FD + CARS + SVM model achieved the best performance for discrimination of nine varieties of sweet maize seeds with classification accuracies of 94.07% and 94.86% for germ up and germ down orientations respectively, which is promising to be a new approach for discrimination the variety of sweet maize seeds.

Rapid and visual detection of the chemical compositions of plant seeds is important but difficult for a traditional seed quality analysis system. In this study, a custom-designed line-scan Raman hyperspectral imaging system was applied for detecting and displaying the main chemical compositions in a heterogeneous maize seed. Raman hyperspectral images collected from the endosperm and embryo of maize seed were acquired and preprocessed by Savitzky-Golay (SG) filter and adaptive iteratively reweighted Penalized Least Squares (airPLS). Three varieties of maize seeds were analyzed, and the characteristics of the spectral and spatial information were extracted from each hyperspectral image. The Raman characteristic peaks, identified at 477, 1443, 1522, 1596 and 1654 cm−1 from 380 to 1800 cm−1 Raman spectra, were related to corn starch, mixture of oil and starch, zeaxanthin, lignin and oil in maize seeds, respectively. Each single-band image corresponding to the characteristic band characterized the spatial distribution of the chemical composition in a seed successfully. The embryo was distinguished from the endosperm by band operation of the single-band images at 477, 1443, and 1596 cm−1 for each variety. Results showed that Raman hyperspectral imaging system could be used for on-line quality control of maize seeds based on the rapid and visual detection of the chemical compositions in maize seeds.

The moisture content (MC) of cucumber seeds was detected nondestructively using two hyperspectral imaging (HSI) systems with complementary spectral ranges. The mean spectrum of each cucumber seed was extracted from hyperspectral images in 400–1000 and 1050–2500 nm separately and it was found that the reflectance spectra decreased as the MC increased in 1050–2500 nm. Calibration models were established by partial least squares regression (PLS) to analyze the predictive ability of preprocessing and wavelength selection methods. The spectra in 400–1000 nm pretreated by Savitzky–Golay smoothing and standard normal variate (SG–SNV) and the 1050–2500 nm spectra pretreated by SG-normalization yielded better results. The optimal wavelengths were obtained by three effective wavelength selection methods, i.e., competitive adaptive reweighted sampling (CARS), iteratively retains informative variables (IRIV), and random frog (RF). Subsequently, the simplified models were built by the selected wavelengths separately. Compared to other developed models, the calibration model established with eight wavelengths chosen by RF from hyperspectral images at 1050–2500 nm achieved optimal performance. The correlation coefficient of prediction (Rpre) was 0.917 and the root mean square error of prediction (RMSEP) was 1.656%. Finally, the visualization of MC distribution was generated at the pixel level. The obtained results in this work indicated that applying HSI technology to measure MC in cucumber seeds was feasible, and the spectrum in 1050–2500 region was more promising than 400–1000 for MC detection. The visualization of MC distribution provided by HSI ensured comprehensive evaluation of MC in single seed level. The selected wavelengths were useful for building a multispectral imaging system to detect MC of cucumber seeds, which could get rid of the seeds with high MC and avoid seed deterioration during storage quickly.

Higher accuracy for a prediction model is the unremitting pursuit in the field of optics nondestructive technology. The multipoint full-transmittance spectra ranging from 650 to 1000 nm were acquired at a speed of 0.5 m/s using an on-line spectrum measurement system. The combination of mean normalization and 11 points smoothing were selected as the best spectral preprocessing method for removing undesirable signal excursion and light scatters existed in the original spectra. By investigating the interference of transmittance spectral intensity on the prediction accuracy with the method of efficient spectrum optimization proposed in our study, we found that those transmittance spectra with intensity lower than 0.4 at 920 nm were inefficient for SSC prediction. Furthermore, three different variable selection algorithms were used to select characteristic band for further optimizing the prediction model, the best prediction model was built based on 45 variables selected by random flog (RF) and the performance of the best model was Rpre = 0.9043 and RMSEP = 0.4787 respectively. All mentioned above illustrated that efficient spectrum optimization method coupled with variable selection algorithms were useful for improving the accuracy and robust of prediction model.