Soluble Solids Content Prediction Model Research Articles

Internal quality evaluation of ‘Fuji’ apples during storage based on bulk optical properties or diffuse reflection and transmission spectra

The purpose of this study is to compare the performance of prediction models based either on the bulk optical properties (BOP) or with models based on the diffuse reflection and transmission spectra to assess the soluble solids content (SSC) or fruit firmness (FF) of ‘Fuji’ apples. Diffuse reflection and transmission spectra of 240 apples in 500–1000 nm were acquired using a self-constructed single integrating sphere system, from which the absorption coefficient (μa) and reduced scattering coefficient (μ′s) were determined. The relationships of μa, μ′s, and the reflection and transmission spectra with the SSC and FF were analyzed, and detection models were established using partial least squares regression coupled with characteristic wavelength selection. Results showed that neither the SSC nor the FF changed obviously during 80-days storage, and the μa and μ′s showed no monotonic increase or decrease pattern. The SSC prediction model based on μa spectra combined with the baseline correction method performed the best (R2p = 0.79, RMSEP = 0.62), while the model based on μ′s spectra was the best for FF (R2p = 0.67, RMSEP = 0.35 N). Overall, the accuracy of the model based on BOP in predicting internal quality was better than that based on reflection and transmission spectra from the same measurement.

Sample selection method using near‐infrared spectral information entropy as similarity criterion for constructing and updating peach firmness and soluble solids content prediction models

AbstractWhen using near‐infrared (NIR) techniques for analysis, model construction and maintenance updates are essential. When model construction is performed in machine learning, the sample set is usually divided into the calibration set and the validation set. The representativeness of the calibration set and the reasonable distribution of the validation set affects the accuracy of the established model. In addition, when maintaining and updating models, selecting the most informative updated sample not only improves the model prediction accuracy but also reduces sample preparation. In this paper, the spectral information entropy (SIE) is proposed to be used as a similarity criterion for dividing the sample set and use this criterion to select updated samples. The Kennard–Stone (KS) and the sample set portioning based on joint x–y distance (SPXY) methods were used for comparison to verify the superiority of the proposed method. The results showed that the model built after dividing the sample set using the SIE method has good prediction effect compared with KS and SPXY method. When predicting soluble solid content (SSC) and hardness, the prediction determination coefficient ( ) was improved by more than 15%, and the root mean square error (RMSE) of prediction was reduced by 50%. In terms of model updating, selecting a small number of updated samples using the SIE method can improve the correlation coefficient ( ) to more than 80%, and updated models' prediction accuracy is higher than that of KS and SPXY method. It is confirmed that the SIE method can make the NIR analysis technique more reliable.

Sorting of navel orange soluble solids content based on online near infrared spectroscopy

Abstract A rapid, nondestructive, and online detection of the internal quality of navel orange cannot only reduce the labor intensity, but also improve the economic benefits of the navel orange. In this paper, an online detection and sorting equipment is designed for navel orange. The transmission spectrum data of 1697 navel oranges are divided into the calibration, prediction, and validation sets, with a ratio of 14:3:3. Pre-processing methods such as first derivative (FD), second derivative (SD), standard normal variate transform (SNV), and multiplicative scatter correction (MSC) were chosen to process the spectra. Accordingly, the soluble solids content prediction model for navel oranges is established using standard normal variable transformation (SNV) and partial least squares (PLS). The determination coefficients (R 2) of the calibration set, prediction set, and validation set are 0.8476, 0.8326, and 0.8025, respectively. Moreover, the corresponding root mean square errors are 0.5097°Brix, 0.5590°Brix, and 0.6048°Brix, respectively. The residual predictive deviation (RPD) value is 2.4510 (i.e., greater than 2.0), indicating that the model performs accurate predictive simulations, and has high reliability. In addition, two classification methods based on the national standard method and the normal probability graph of the soluble solids content of navel oranges were used to classify navel oranges into three classes for online validation. 185 navel oranges were selected for online validation, in which the classification method based on the normal probability graph of the soluble solids content of navel oranges was more effective and its average sorting accuracy was 81.13 %. Likewise, the mean absolute error (MAE) is 0.4613°Brix. The experimental results show that the online sorting equipment possesses high sorting accuracy and can be practically used for actual postharvest processing.

Research on an Improved Non-Destructive Detection Method for the Soluble Solids Content in Bunch-Harvested Grapes Based on Deep Learning and Hyperspectral Imaging

The soluble solids content (SSC) is one of the important evaluation indicators for the internal quality of fresh grapes. However, the current non-destructive detection method based on hyperspectral imaging (HSI) relies on manual operation and is relatively cumbersome, making it difficult to achieve automatic detection in batches. Therefore, in this study, we aimed to conduct research on an improved non-destructive detection method for the SSC of bunch-harvested grapes. This study took the Shine-Muscat grape as the research object. Using Mask R-CNN to establish a grape image segmentation model based on deep learning (DL) applied to near-infrared hyperspectral images (400~1000 nm), 35 characteristic wavelengths were selected using Monte Carlo Uninformative Variable Elimination (MCUVE) to establish a prediction model for SSC. Based on the two abovementioned models, the improved non-destructive detection method for the SSC of bunch-harvested grapes was validated. The comprehensive evaluation index F1 of the image segmentation model was 95.34%. The Rm2 and RMSEM of the SSC prediction model were 0.8705 and 0.5696 Brix%, respectively, while the Rp2 and RMSEP were 0.8755 and 0.9177 Brix%, respectively. The non-destructive detection speed of the improved method was 16.6 times that of the existing method. These results prove that the improved non-destructive detection method for the SSC of bunch-harvested grapes based on DL and HSI is feasible and efficient.

Fusion models for detection of soluble solids content in mandarin by Vis/NIR transmission spectroscopy combined external factors

In the process of acquiring the visible/near infrared (Vis/NIR) spectra of mandarin to predict the soluble solids content (SSC), spectra information is often influenced by the physical property of citrus. Thereupon affects the accuracy and robustness of the SSC prediction model. Calibration models were established by partial least squares (PLS) regression methods based on three spectral measurement orientations (Mandarin calyx–stem axis vertical with stem upward (A); Mandarin calyx–stem axis vertical with stem downward (B); Mandarin calyx–stem axis horizontal (C)) respectively. Novel methods that considering mandarin external factors (diameter, color and peel thickness) were established. The methods were associated with wavelength selection algorithms and their combination with alternative methods to build multivariate regression models. After evaluating the models, the model performance was improved by the external factor inclusion methods, among which the color and diameter united inclusion method combined with ACO-PLS algorithm performed the best, with Rp = 0.987 and RMSEP = 0.19 °Brix for its prediction model. This study serves as a reference and theoretical support for determining the influencing factors that affect fruit internal quality determination using near infrared spectroscopy.

Study on the Effect of Apple Size Difference on Soluble Solids Content Model Based on Near-Infrared (NIR) Spectroscopy

Soluble solids content (SSC) is a vital evaluation index for the internal quality of apples, and NIR spectroscopy is the preferred technique for predicting the SSC of apples. Due to the differences in fruits’ sizes, their SSC prediction models have poor robustness and low prediction accuracy, so it is important to eliminate the effects brought by the differences in fruit sizes to improve the accuracy of fruit sorting models. The NIR spectra of apples with different fruit sizes were collected by applying NIR spectroscopy online detection device, and after various preprocessing of the spectra, the partial least squares (PLS) models of apple SSC were established, respectively, and then the modeling set in the apple fruit size group of 75 mm–85 mm was used to predict the prediction set samples in the apple fruit size group of 65 mm–75 mm and 85 mm–95 mm, respectively. To better address the effects of apple size differences, data fusion techniques were used to perform an intermediate fusion of apple fruit diameter and spectra, firstly, the competitive adaptive reweighting algorithm (CARS) and the continuous projection algorithm (SPA) were used to select spectral variables and build their prediction models for apple SSC, respectively, and the results showed that the models built with 61 spectral variables selected by CARS had better performance, greatly reduced the amount of data involved in modeling, effectively simplified the model, and improved the stability of the model. The apple size variables were added to the wavelength variables selected by CARS, and the data were normalized to establish a PLS model of apple SSC with the normalized spectral and apple fruit diameter data, and the results showed that the size compensation model based on intermediate fusion had the best prediction performance, with the prediction set Rp of 0.886 for fruit diameter of 65 mm–75 mm, RMSEP of 0.536%, and its prediction set Rp was 0.913 and RMSEP was 0.497% for the fruit diameter of 85 mm–95 mm. Therefore, adding the fruit diameter variable to establish the size-compensated model of apple SSC can improve the prediction performance of the model.

Effects of Orientations and Regions on Performance of Online Soluble Solids Content Prediction Models Based on Near-Infrared Spectroscopy for Peaches

Predicting the soluble solid content (SSC) of peaches based on visible/near infrared spectroscopy has attracted widespread attention. Due to the anisotropic structure of peach fruit, spectra collected from different orientations and regions of peach fruit will bring variations in the performance of SSC prediction models. In this study, the effects of spectra collection orientations and regions on online SSC prediction models for peaches were investigated. Full transmittance spectra were collected in two orientations: stem-calyx axis vertical (Orientation1) and stem-calyx axis horizontal (Orientation2). A partial least squares (PLS) method was used to evaluate the spectra collected in the two orientations. Then, each peach fruit was divided into three parts. PLS was used to evaluate the corresponding spectra of combinations of these three parts. Finally, effective wavelengths were selected using the successive projections algorithm (SPA) and competitive adaptive reweighted sampling (CARS). Both orientations were ideal for spectra acquisition. Regions without peach pit were ideal for modeling, and the effective wavelengths selected by the SPA led to better performance. The correlation coefficient and root mean square error of validation of the optimal models were 0.90 and 0.65%, respectively, indicating that the optimal model has potential for online prediction of peach SSC.

Nondestructive determination of SSC in an apple by using a portable near-infrared spectroscopy system

The soluble solids content (SSC) is an important factor in the internal quality detection of apples. It is essential to have reliable and high-speed measurement of the SSC. However, almost all traditional equipment is inconvenient and expensive. We designed a handheld nondestructive SSC detector based on near-infrared (NIR) spectroscopy, which is composed of a portable NIR spectrometer, cloud server, smartphone app, and prediction model of SSC. We preprocessed the spectrum with multiplicative scatter correction (MSC), standard normal variable transformation (SNV), and Savitzky-Golay (S-G) smoothing algorithms. Besides, the linear weight reduction of the particle swarm optimization algorithm is carried out, and we establish the model of an extreme learning machine optimized with the improved particle swarm optimization (IPSO-ELM) algorithm. The R2, root mean square error of prediction (RMSEP), and residual prediction deviation (RPD) of the model are 0.993, 0.0155, and 11.6, respectively, which are better than the traditional model obviously. In addition, the number of wavelengths reduced from 228 to 70 as the model is simplified with the uninformative variable elimination (UVE) method. The time of training is reduced by 29.30% compared with the original spectrum. It can be verified that the IPSO-ELM model has good prediction performance, and the NIR diffuse reflectance spectroscopy is a reliable nondestructive measurement of SSC in apples.

Online detection of soluble solids content and maturity of tomatoes using Vis/NIR full transmittance spectra

To meet the demand for high-quality tomatoes of different commercial purposes, the optimal models for predicting soluble solids content (SSC)and classifying maturities of tomatoes were developed simultaneously based on the Vis-NIR spectra collected by an online full transmittance spectrum measurement system. The intensity of full transmittance spectra at maturities of green, turning, pink, light red and red increased with their mature stages, indicating that the optical properties of tomatoes correlated significantly with their growth stages. Hence, the full transmittance has the potential to distinguish the maturities of tomatoes. The prediction models of SSC and classification models of maturity were built by partial least squares (PLS) regression and extreme learning machine (ELM) classifier based on the spectral data obtained by different detection orientations and different preprocessing methods, respectively. The results showed that detection orientations significantly affected the performance of SSC prediction models and maturity classification models, and the best model for SSC prediction was developed using the full wavelength spectra preprocessed by multiplicative scatter correction (MSC) with T1 orientation (the stem-top axis is vertical to conveyor belt). For samples in the prediction set, the correlation coefficient (Rp) and root mean square error (RMSEP) and residual predictive deviation (RPD) of the best SSC prediction model were 0.75, 0.27 °Brix and 1.49, respectively. The best classification model for five classes was built based on the first ten principle components (PC) of spectra preprocessed by MSC with T1 orientation. The success rate of model for classifying the total, green, turning, pink, light red and red stages were 80.08%, 94.62%, 77.31%, 62.69%, 78.85%, 89.62%, respectively. To further investigate the classification ability of ELM classifier and simplify the classification complexity of maturities, the initial five maturities were converted into three maturities, namely immature, intermediate (turning ​+ ​pink) and mature (light red ​+ ​red). Compared with the classification model of five classes, model that was established for classifying three types of maturities obtained better results with accuracies of 91.31%, 95.38%, 89.23% and 91.35% for total, immature, intermediate and mature stages, respectively, based on the full wavelength spectra pretreated by MSC with T1 orientation. Overall, it was optimal for detection models of SSC and maturity by combining T1 orientation with MSC preprocessing, and the full transmittance mode used in this study has further potential applications for quality evaluation of agriculture products.

Neural Network Based Prediction of Soluble Solids Concentrationin Oriental Melon Using VIS/NIR Spectroscopy

Highlights Non-destructive soluble solids content prediction model for oriental melon was developed based on NIR spectrum data. Not only the classical ML or Neural-Network methods, but also the mixture of both techniques have also been tried. Comparing the various pre-processing methods, the MSC-PLS-ANN model showed the best results. MSC-PLS-ANN model demonstrated 6% of improvement in RMSE score over the PLSR model, which is commonly used in commercial products Abstract. Models for predicting the soluble solids concentration (SSC) of oriental melons were developed and evaluated by applying near infrared spectroscopy and an artificial neural network technique. For the evaluation, a total of 300 oriental melons, both ripe and unripe, were mixed together and sampled. To develop an SSC prediction model, the actual SSC values of specimens having the same spectra as those of the visible/near infrared wavelength bands were measured. The measured spectra were preprocessed using eight methods [Multiplicative Scatter Correction (MSC), Standard Normal Variate (SNV), Robust Normal Variate, Savitzky-Golay 1st and 2nd; Min-Max Normalization; Robust Normalization; Standardization], and the SSC prediction model was developed by applying three techniques (Partial Least Squared Regression [PLSR], Artificial Neural Network [ANN], and Convolutional Neural Network [CNN]). Among them, the PLSR technique also applied a Variable Importance in Projection (VIP) method for wavelength selection. Among the PLSR-based SSC prediction models, the SNV-preprocessed PLSR model showed the best SSC prediction performance (RMSEtest, 0.67; R2test, 0.81). Among the ANN-based models, the MSC-preprocessed PLS-ANN model showed the best SSC prediction performance (RMSEtest: 0.63, R2test: 0.83). Among the CNN-based models, the DeepSpectra model was applied, but showed the lowest prediction performance (RMSEtest: 0.79, R2test: 0.74). In conclusion, among the three SSC prediction algorithms tested in this study, the PLS-ANN-based prediction model showed the best SSC prediction performance, which was found to be higher than that of the PLSR-based SSC prediction model applied to the sugar sorters currently used in agricultural products at processing centers. Keywords: Artificial Neural Network, Convolution Neural Network, Korean melon, VIP-PLSR, VIS/NIR spectroscopy.

基于可见光和近红外光谱的便携式苹果可溶性固形物含量检测仪的研制

In order to detect the soluble solids content of apples quickly and accurately, a portable apple soluble solids content detector based on USB2000 + micro spectrometer was developed. The instrument can communicate with computer terminal and mobile app through network port, Bluetooth and other ways, which can realize the rapid acquisition of apple spectral information. Firstly, the visible / near-infrared spectrum data and soluble solids content information of 160 apple samples were collected; secondly, the spectral data preprocessing methods were compared, and the results showed that the prediction model of sugar content based on partial least square (PLS) method after average smoothing preprocessing was accurate. The correlation coefficient (RP) and root mean square error (RMSEP) of the prediction model were 0.902 and 0.589 ° Brix, respectively. Finally, on the basis of average smoothing preprocessing, competitive adaptive reweighted sampling (CARS) and successive projections algorithm (SPA) were used to optimize the wavelength of spectral data, and PLS model was constructed based on the selected 17 characteristic wavelengths, which can increase the accuracy of soluble solids content prediction model, increase the RP to 0.912, and reduce RMSEP to 0.511 ° Brix. The portable visible / near infrared spectrum soluble solids prediction model based on the instrument and method has high accuracy, and the detector can quickly and accurately measure the soluble solids content of apple.

Optimization of soluble solids content prediction models in ‘Hami’ melons by means of Vis-NIR spectroscopy and chemometric tools

Nondestructive assessment of internal quality in fruit with thick rind and large size was always a challenge in terms of the fruit quality automatic grading application. This study compared different spectral measurement positions and prediction models for determining soluble solids content (SSC) of ‘Hami’ melon. Visible and short wavelength near-infrared with the spectral range of 550–950 nm was used and three local positions (calyx, equator and stem regions) in each sample were selected for spectral collection, respectively. The partial least squares (PLS) was employed to establish three local calibration models (calyx-model, equator-model and stem-model) and calyx-model was proved to be optimal one for SSC prediction of ‘Hami’ melon. In order to optimize the calyx-model, four algorithms including successive projections algorithm (SPA), Monte Carlo-uninformative variable elimination (MC-UVE), competitive adaptive reweighted sampling (CARS) and combination algorithm MC-UVE-SPA were applied to select effective variables from full spectra, respectively. And, the non-linear least square support vector machine (LSSVM), linear multiple linear regression (MLR) and PLS models were also developed based on those selected effective variables, respectively. The results indicated that MC-UVE-SPA was the optimal method for selecting the effective variable. Based on 18 variables selected by MC-UVE-SPA, all established models including MC-UVE-SPA-PLS, MC-UVE-SPA-LSSVM and MC-UVE-SPA-MLR can successfully predict the SSC of all ‘Hami’ melon fruit samples with the prediction accuracy of RMSEP = 0.95–0.99 °Brix.

A multi-region combined model for non-destructive prediction of soluble solids content in apple, based on brightness grade segmentation of hyperspectral imaging

On-line estimation of fruit internal attributes based on visible-near infrared spectrum is an effective approach for improving fruit value and farmer income. Brightness correction of the hyperspectral image can significantly affect the accuracy of quality estimation in fruits with spherical geometric structure, while pixel by pixel correction method is time-consuming and impractical in on-line rapid detection. To improve the flexibility and speed of the soluble solids content (SSC) estimation model of intact apple, the spectra of core region, middle region and outer region were extracted from hyperspectral reflectance imaging over the region of 400–1000 nm with a brightness grade segmentation method, and then a multi-region combined partial least square (MCPLS) prediction model was built. Results showed that MCPLS method achieved better results than traditional PLS and multi-region average PLS methods. To further improve the applicability of the prediction model in practice, 21 wavelengths effective for SSC estimation were selected by successive projection algorithm and used to rebuild the prediction model using MCPLS method; the correlation coefficient, root mean square error of prediction set and residual predictive deviation were 0.9132, 0.3929 and 2.1652 respectively. Additionally, multi-region combined method just needs to compute the average spectra of each region, which significantly improved the detection speed by comparison with previous pixel by pixel brightness correction method. Hence the multi-region combined prediction model of SSC was developed based on the spectral contribution of each region to SSC prediction and the method of brightness grade segmentation.

Comparative Study of Two Different Strategies for Determination of Soluble Solids Content of Apples From Multiple Geographical Regions by Using FT-NIR Spectroscopy

Apple is one of the most popular fresh fruits with an extensive scope of regions owing to its nutrition and sweet in flavor. There is a large difference in the composition of the fruits growing in varying regions because of the variation in the growing regions, such as temperature, soil nutrients, etc. As a result, it is of significance to decrease the impact of region variability on the measurement of soluble solids content (SSC) in apples. To lessen the impact of region variability and enhance the predictive ability of on the model, our manuscript compared the performance of the two multi-region prediction models for the estimation of SSC in apples from multiple geographical regions. One multi-region prediction model was developed by merging SSC values and spectral data of all samples from multiple regions. The other multi-region prediction model was built for the determination of SSC in combination with region discriminant, model search strategy, and single-region models. Support vector machine (SVM) was applied to establish the model for discriminating the apples from multiple geographical regions. It was found that the region discriminant model achieved great results, with the classification accuracy of 99.52%. By comparing and analyzing the two multi-region prediction models, the optimal multi-region prediction model was obtained. Finally, to decrease the irrelevant spectral information and reduce the computational cost, the multi-region SSC prediction model was optimized in combination with various spectral preprocessing methods (multiple scatter correction (MSC), standard normal variate (SNV), and first derivative (FD) correction) and variable selection methods (Monte Carlo uninformative variables elimination (MC-UVE), competitive adaptive reweighted sampling (CARS), and random frog (RF)). The overall results denoted that it was more accurate to estimate SSC in apples from the different geographical regions by using the multi-region models based on the region discriminant model in combination with SNV preprocessing algorithm and MC-UVE variable selection algorithm, and the prediction accuracy preceded the single-region models.

Nondestructive determination of soluble solids content of persimmons by using dielectric spectroscopy

ABSTRACTTo explore the feasibility of dielectric spectroscopy in predicting soluble solids content (SSC) of persimmons during postharvest storage period, the dielectric constant spectra and dielectric loss factor spectra of 105 ‘Shui’ persimmons were measured from 20 MHz to 4500 MHz. Based on the joint x-y distances algorithm, the persimmon samples were divided into two sets: 70 samples in calibration set and 35 samples in prediction set. One hundred and seventy-four, 14, and 24 variables were extracted as characteristic variables from full dielectric spectra (FS) by uninformative variables elimination (UVE), successive projection algorithm (SPA), and competitive adaptive reweighted sampling (CARS), respectively. Partial least squares (PLS) and least squares support vector machine (LSSVM) were applied to build SSC prediction models using FS and characteristic variables extracted by UVE, SPA, and CARS. The results indicated that LSSVM models offered better performance than PLS models at same input variables. CARS-LSSVM had the best SSC determination performance with the correlation coefficient and root-mean-square error of prediction set of 0.970 and 0.494°Brix. This study indicates that dielectric spectroscopy technique combined with characteristic variables selection methods is promising for determining SSC of persimmons.

Predicting Apple Firmness and Soluble Solids Content Based on Hyperspectral Scattering Imaging Using Fourier Series Expansion

Abstract. This article reports on using a Fourier series expansion method to extract features from hyperspectral scattering profiles for apple fruit firmness and soluble solids content (SSC) prediction. Hyperspectral scattering images of ‘Golden Delicious’ (GD), ‘Jonagold’ (JG), and ‘Delicious’ (RD) apples, harvested in 2009 and 2010, were acquired using an online hyperspectral imaging system over the wavelength region of 500 to 1000 nm. The moment method and Fourier series expansion method were used to analyze the scattering profiles of apples. The zeroth-first order moment (Z-FOM) spectra and Fourier coefficients were extracted from each apple, which were then used for developing fruit firmness and SSC prediction models using partial least squares (PLS) and least squares support vector machine (LSSVM). The PLS models based on the Fourier coefficients improved the standard errors of prediction (SEP) by 4.8% to 19.9% for firmness and by 2.4% to 13.5% for SSC, compared with the PLS models using the Z-FOM spectra. The LSSVM models for the prediction set of Fourier coefficients achieved better SEP results, with improvements of 4.4% to 11.3% for firmness and 2.8% to 16.5% for SSC over the LSSVM models for the Z-FOM spectra data and 3.7% to 12.6% for firmness and 5.4% to 8.6% for SSC over the PLS models for the Fourier coefficients. Experiments showed that Fourier series expansion provides a simple, fast, and effective means for improving Keywords: Apples, Firmness, Fourier series expansion, Hyperspectral scattering imaging, Least squares support vector machine, Partial least squares, Soluble solids content.

Nondestructive Determination of Soluble Solids Content of ‘Fuji’ Apples Produced in Different Areas and Bagged with Different Materials During Ripening

To investigate the feasibility of hyperspectral imaging technique in nondestructive determination of soluble solids content (SSC) of fruits produced in different places and bagged with different materials during ripening, the near infrared hyperspectral reflectance images were acquired on 196 ‘Fuji’ apples picked from four orchards in different areas and bagged with polyethylene film or light-impermeable paper. Mean reflectance spectrum from the regions of interest in the hyperspectral image of each apple was extracted. Standard normal variate (SNV) was used to eliminate the effect of instrument and environment on spectra. The sample set partitioning based on joint x–y distances method was applied to divide the samples into calibration set and prediction set as the ratio of 3:1. Successive projection algorithm (SPA) and uninformative variable elimination (UVE) method were used to select effective wavelengths (EWs) from the full spectra. Partial least squares (PLS), least squares support vector machine (LSSVM), and extreme learning machine (ELM) were used to develop SSC determination models. The results showed that 24 and 122 EWs were selected by SPA and UVE, respectively. The selection of EWs was helpful to SSC determination performance improvement. The optimal SSC prediction model was LSSVM based on selected EWs by SPA, with the correlation coefficient and root-mean-square error of prediction set of 0.878 and 0.908 °Brix, respectively. This study indicates that hyperspectral imaging technique could be used to determine SSC of intact apples produced in different places and bagged with different materials during ripening.

Prediction of Soluble Solids Content of Jujube Fruit Using Hyperspectral Reflectance Imaging

To predict soluble solids content (SSC) of jujube fruits, a hyperspectral imaging technique has been used for acquiring reflectance images from 200 samples in the spectral regions of 900-1700nm. Hyperspectral images of jujubes were evaluated from the regions of interest using principal component analysis (PCA) with the goal of selecting five optimal wavelengths (1034, 1109，1231，1291 and 1461nm). Prediction model of SSC (Rp=0.9027, RMSEP=1.9845) were built based on BP neural network. This research has demonstrated the feasibility of implementing hyperspectral imaging technique for sorting jujube fruit for SSC to enhance the product quality and marketability.

Optimal Wavelength Selection for Hyperspectral Scattering Prediction of Apple Firmness and Soluble Solids Content

Hyperspectral scattering is a promising technique for nondestructive quality measurement of apple fruit, and extraction of the most useful information from the hyperspectral scattering data is critical for accurate assessment of fruit firmness and soluble solids content (SSC). In this article, a hierarchical evolutionary algorithm (HEA) approach coupled with subspace decomposition and partial least squares regression is proposed to select the optimal wavelengths from hyperspectral scattering profiles of 'Golden Delicious' apples for predicting fruit firmness and SSC. Six hundred apples were tested in the experiment, 400 of which were used for calibration and the remaining 200 apples for validation. Seventeen optimal wavelengths were selected for firmness prediction, which nearly spanned the entire spectral range of 500 to 1000 nm, and 16 optimal wavelengths, all of which were above 600 nm, were selected in the SSC prediction model. The model using the 17 optimal wavelengths for predicting firmness yielded better results (r = 0.857, root mean square error of prediction or RMSEP = 6.2 N) than the full spectrum model (r = 0.848, RMSEP = 6.4 N). For predicting SSC, the model using the 16 optimal wavelengths also yielded better results (r = 0.822, RMSEP = 0.78%) than the full spectrum model (r = 0.802, RMSEP = 0.83%). The HEA approach provided an effective means for optimal wavelength selection and improved the prediction of firmness and SSC in apples compared with the approach using the full spectrum.

Vis/NIR estimation at harvest of pre- and post-storage quality indices for ‘Royal Gala’ apple

Visible/Near Infrared (Vis/NIR) spectrometric measurements, made with a fast low cost polychromatic spectrometer operating over the range 500–1100 nm, have been made on ‘Royal Gala’ apples (Malus domestica Borkh.). The spectral measurements were taken on fruit from eight orchards in two growing regions and in each of two seasons through a 7-week period centred around the commercial harvesting period. Data analysis involved creating predictive models, by chemometric modelling, for various quality indices measured both immediately after harvest and after 6 weeks cool storage. The quality indices included background colour (BC), starch pattern index (SPI), soluble solids content (SSC), penetrometer firmness, quantitative starch (QS) and titratable acidity (TA). The predictive models were significant in regression terms, typically explaining between 50 and 80% of the data set variance, but they were, nonetheless, very poor in prediction terms. For each quality index, the ratio of data standard deviation to the root mean square error of prediction (RMSEP) was always well less than 3; a ratio value considered the minimum for only coarse screening/grading of fruit. Most of the prediction models appear to be primarily dependent on changes in the chlorophyll absorbance peak, which dramatically reduces in intensity during the progression of the fruit through the harvest period, rather than on the constituent or property of direct interest. Exceptions were the SSC prediction models that are dependent, to varying degrees, on the relevant constituents of interest (soluble carbohydrates). The SSC predictions were much better post-storage than at harvest time (RMSEP ∼0.50 and 0.72%, respectively) indicating, we theorise, the confounding influence of starch, present in the fruit at harvest time, on the SSC model.