Data fusion of spectral,thermal and canopy height parameters for improved yield prediction of drought stressed spring barley

Yield modelling based on visible and near infrared spectral information is extensively used in proximal and remote sensing for yield prediction of crops. Distance and thermal information contain independent information on canopy growth, plant structure and the physiological status. In a four-years′ study hyperspectral, distance and thermal high-throughput measurements were obtained from different sets of drought stressed spring barley cultivars. All possible binary, normalized spectral indices as well as thirteen spectral indices found by others to be related to biomass, tissue chlorophyll content, water status or chlorophyll fluorescence were calculated from hyperspectral data and tested for their correlation with grain yield. Data were analysed by multiple linear regression and partial least square regression models, that were calibrated and cross-validated for yield prediction. Overall partial least square models improved yield prediction (R² = 0.57; RMSEC = 0.63) compared to multiple linear regression models (R² = 0.46; RMSEC = 0.74) in the model calibration. In cross-validation, both methods yielded similar results (PLSR: R² = 0.41, RMSEV = 0.74; MLR: R² = 0.40, RMSEV = 0.78). The spectral indices R₇₈₀/R₅₅₀, R₇₆₀/R₇₃₀, R₇₈₀/R₇₀₀, the spectral water index R₉₀₀/R₉₇₀ and laser and ultrasonic distance parameters contributed favourably to grain yield prediction, whereas the thermal based crop water stress index and the red edge inflection point contributed little to the improvement of yield models. Using only more uniform modern cultivars decreased the model performance compared to calibrations done with a set of more diverse cultivars. The partial least square models based on data fusion improved yield prediction (R² = 0.62; RMSEC = 0.59) compared to the partial least square models based only on hyperspectral data (R² = 0.48; RMSEC = 0.69) in the model calibration. This improvement was confirmed by cross-validation (data fusion: R² = 0.39, RMSEV = 0.76; hyperspectral data only: R² = 0.32, RMSEV = 0.79). Thus, a combination of spectral multiband and distance sensing improved the performance in yield prediction compared to using only hyperspectral sensing.     

Bruise susceptibility of banana peel in relation to genotype and post-climacteric storage conditions

The aim of this study was to understand the genotypic factors and post-climacteric storage conditions that affect bruise susceptibility of banana peel. Putative physicochemical indicators of bruise susceptibility, including peel electrolyte leakage (PEL), total polyphenolic content, hardness, water content, and peel thickness, were investigated. Bruise susceptibility is the lowest impact energy needed to produce visible bruising by an object dropped on post-climacteric banana fruit from a pre-determined height, converted into impact energy (20–200 mJ with a 20 mJ increment). The bananas were stored either at 18 °C throughout ripening or at 13 °C between the 2nd and 6th day after ethylene induction. Five cultivars with contrasting susceptibility to impact bruises were used. Neither Grande Naine nor hybrid Flhorban925 bruised at the maximum impact energy (200 mJ) during ripening whatever the storage conditions. A gradient in bruise susceptibility was observed among the other cultivars: French Corne > Fougamou > hybrid Flhorban916. Bruise susceptibility increased during ripening and was higher in bananas stored at 18 °C. The lower ripening temperature resulted in a two-day delay to fruit maturity as well as in bruise susceptibility. Bruise susceptibility was positively correlated with PEL (R = 0.78) and to a lesser extent negatively correlated with hardness (R = −0.45), and was not correlated with polyphenolic content. In conclusion, membrane permeability provides the first clue to understanding bruise susceptibility.     

Examination of the quality of spinach leaves using hyperspectral imaging

The present research is focused on the application of hyperspectral images for the supervision of quality deterioration in ready to use leafy spinach during storage (Spinacia oleracea). Two sets of samples of packed leafy spinach were considered: (a) a first set of samples was stored at 20 °C (E-20) in order to accelerate the degradation process, and these samples were measured the day of reception in the laboratory and after 2 days of storage; (b) a second set of samples was kept at 10 °C (E-10), and the measurements were taken throughout storage, beginning the day of reception and repeating the acquisition of Images 3, 6 and 9 days later. Twenty leaves per test were analyzed. Hyperspectral images were acquired with a push-broom CCD camera equipped with a spectrograph VNIR (400–1000 nm). Calibration set of spectra was extracted from E-20 samples, containing three classes of degradation: class A (optimal quality), class B and class C (maximum deterioration). Reference average spectra were defined for each class. Three models, computed on the calibration set, with a decreasing degree of complexity were compared, according to their ability for segregating leaves at different quality stages (fresh, with incipient and non-visible symptoms of degradation, and degraded): spectral angle mapper distance (SAM), partial least squares discriminant analysis models (PLS-DA), and a non linear index (Leafy Vegetable Evolution, LEVE) combining five wavelengths were included among the previously selected by CovSel procedure. In sets E-10 and E-20, artificial images of the membership degree according to the distance of each pixel to the reference classes, were computed assigning each pixel to the closest reference class. The three methods were able to show the degradation of the leaves with storage time.     

Analysis of the main secondary metabolites produced in tomato (Lycopersicon esculentum,Mill.) epicarp tissue during fruit ripening using fluorescence techniques

Tomato (Lycopersicon esculentum L.) fruit are an important source of antioxidant (mainly pigment) compounds, as well as lycopene, β-carotene, ascorbic acid and polyphenols. Differentiation of the final product in the market requires an accurate evaluation of these value-adding compounds. Because of this, we have undertaken a comparison of the spectral characterisation of the tomato fruit surface pigments from the immature to over-ripe stage, using spectroscopy techniques based on visible fluorescence emission upon excitation in the same or ultraviolet spectral regions. The aim was to verify the spectral band for optimal conditions for fruit harvesting using non-destructive techniques. The pattern of pigment composition changed markedly during ripening and showed progressive disappearance of chlorophyll with a concomitant increase in carotenoids until the fully ripe stage. The main fluorescence spectral features belonging to anthocyanins, flavonoids, carotenoids and chlorophyll a after excitation of skin tomato pigments at different laser wavelengths was identified. In comparing, the fluorescence spectral ratios at the excitation wavelength λ_exc = 266 nm, significant differences were obtained for the spectral ratios of chlorophyll/flavonoids and carotenoids/chlorophyll. Positive correlation coefficients were found for the carotenoids/flavonoids (0.780) ratios and negative ones for the carotenoids/chlorophyll ratios (−0.513).Analysis of fluorescence resulted in determination of the most useful laser radiation for remote non-invasive measurements with laser-induced fluoresence (LIF): for the ripening stage, λ_exc = 266 nm was the optimal laser wavelength, since the induced fluorescence spectra obtained appeared to differ with the physiological stage of the fruit.     

Table olive cultivar susceptibility to impact bruising

Developing mechanical harvesting for table olives will require decreasing fruit damage during harvest and postharvest handling, transport and storage. The susceptibility to bruising and its development over time were studied in three table olive varieties, cv. ‘Manzanilla’, ‘Gordal Sevillana’ and ‘Hojiblanca’. Bruising was produced with controlled energy impacts of 56, 26, 13 mJ. A strong correlation (r² = 0.77–0.90) between bruise volume and impact energy was demonstrated. Bruise susceptibility was higher in the Manzanilla variety, followed by Hojiblanca and Gordal Sevillana cultivars. Bruise time evolution was evaluated using a spectrophotometer for visible and near infrared regions. A bruise index was developed using different wavelengths, 545, 670 and 800 nm. Most darkening due to the browning process happened within 1 h, was exponential and dependent on impact energy level. The discoloration was greatest in the Manzanilla, followed by Hojiblanca and Gordal Sevillana olives.     

Detection of fruit fly infestation in pickling cucumbers using a hyperspectral reflectance/transmittance imaging system

Fruit fly infestation can be a serious problem in pickling cucumber production. In the United States and many other countries, there is zero tolerance for fruit flies in pickled cucumber products. Currently, processors rely on manual inspection to detect and remove fruit fly-infested cucumbers, which is labor intensive and also prone to error due to human fatigue and the difficulty of visually detecting infestation that is hidden inside the fruit. In this research, a laboratory hyperspectral imaging system operated in an integrated mode of reflectance and transmittance was used to detect fruit fly-infested pickling cucumbers. Hyperspectral reflectance (450–740 nm) and transmittance (740–1000 nm) images were acquired simultaneously for 329 normal (infestation-free) and fruit fly-infested pickling cucumbers of three size classes with the mean diameters of 16.8, 22.1, and 27.6 mm, respectively. Mean spectra were extracted from the hyperspectral image of each cucumber, and they were then corrected for the fruit size effect using a diameter correction equation. Partial least squares discriminant analyses for the reflectance, transmittance and their combined data were performed for differentiating normal and infested pickling cucumbers. With reflectance mode, the overall classification accuracies for the three size classes and the mixed class were between 82% and 88%, whereas transmittance achieved better classification results with the overall accuracies of 88–93%. Integration of reflectance and transmittance did not result in noticeable improvements, compared to transmittance mode. The hyperspectral imaging system performed better than manual inspection, which had an overall accuracy of 75% and whose performance decreased significantly for smaller size cucumbers. This research demonstrated that hyperspectral imaging is potentially useful for detecting fruit fly-infested pickling cucumbers.     

Non-destructive prediction of translucent flesh disorder in intact mangosteen by short wavelength near infrared spectroscopy

A non-destructive measurement and data evaluation technique to predict an internal translucent flesh disorder in intact mangosteen fruit is proposed by using short wavelength near infrared (SW-NIR) transmittance spectroscopy. The optimum conditions of measurement were investigated for spectra acquisition at an integration time of 78 ms with a 200 W light source. The NIR absorption spectra of 193 mangosteen samples were obtained in the wavelength range from 640 to 980 nm on four sides of each sample. The best result from a discriminant analysis for leave-one-out cross-validation was 92.0% classification accuracy. The results showed that the hardening pericarp disorder influenced the accuracy of the classification. This study demonstrates that SW-NIR spectroscopy can be used to accurately predict translucent flesh disorder in intact mangosteens.     

NIR spectroscopy for the optimization of postharvest apple management

Apples can be stored for long time under controlled temperature and atmosphere conditions, and therefore, non-destructive and rapid tools are required to assess fruit quality and to monitor changes during the postharvest period. The aim of this study was to evaluate the feasibility of NIR spectroscopy to optimize postharvest apple management and to follow changes in fruit quality during storage. An FT-NIR system operating in diffuse reflectance in the range 12,500–3600 cm⁻¹ was used to evaluate the physico-chemical (dry matter, soluble solids, colour and firmness) and some nutraceutical characteristics (total phenolics, total flavonoids and antioxidant activity) of ‘Golden Delicious’ apples, which were stored for about six months at 1 °C in controlled atmosphere, over two subsequent years. Spectral data were elaborated by PLS regression and LDA classification techniques. Good correlation models between spectral data and chemical and physical parameters were obtained for soluble solids, a* colour coordinate and firmness (0.81 < R² < 0.90 in calibration and 0.79 < R² < 0.89 in cross validation). Even higher correlation values (0.89 < R² < 0.95 in calibration and 0.86 < R² < 0.92 in cross validation) were obtained for indexes correlated to the antioxidant capacity of apples. The classification technique Linear Discriminant Analysis was applied to spectral data, in order to discriminate apples on the basis of storage time. Average correct classification was higher than 93% in validation and close to 100% in calibration, indicating high potential of NIR spectroscopy for the estimation of storage time of apple lots.     

Spectral reflectance characteristics of citrus canker and other peel conditions of grapefruit

The presence of one fruit infected with citrus canker in a shipment may render the whole shipment unmarketable. Therefore it is important to classify fruit infected with citrus canker in the packinghouse before shipping the produce. The purpose of this research was to determine the significant wavelengths that could be used to classify canker among other peel conditions using grapefruit, a variety susceptible to canker. A spectrophotometer, with a wavelength range of 200–2500 nm, was used to measure the spectral reflectance data of peels from market quality fruit and fruits that were infected with cake melanose, wind scar, copper burn, greasy spot and canker. Statistical data analysis was performed on the spectral reflectance data to identify the wavelengths that had maximum discriminatory potential among the different peel conditions and to derive a discriminant function from the identified wavelengths. The discriminatory wavelengths were identified in the visible and the visible near infrared range. In addition, the classification based on the derived discriminant function resulted in 100% classification of canker. These results provide fundamental and practical use in the development of an automatic fruit sorter for canker classification based on spectral reflectance. Future research would involve the development of a vision-based classification system using the significant wavelengths identified in this study.     

‘Rocha’ pear firmness predicted by a Vis/NIR segmented model

We present a segmented partial least squares (PLS) prediction model for firmness of ‘Rocha’ pear (Pyrus communis L.) during fruit ripening under shelf-life conditions. Pears were collected from three different orchards. Orchard I provided the pears for model calibration and internal validation (set 1). These were transferred to shelf-life in the dark at 20 ± 2 °C and 70% RH, immediately after harvest. External validation was performed on the pears from the other two orchards (sets 2 and 3), which were stored under different conditions before shelf-life. Fruit was followed in the shelf-life period by visible/near infrared reflectance spectroscopy (Vis/NIRS) in the range 400–950 nm. The correlation between firmness and the reflectance at some wavelength bands was markedly different depending on ripening stage. A segmented partial least squares model was then constructed to predict firmness. This PLS model has two segments: (1) unripe and ripening/ripe pears (high firmness); (2) over-ripe pears (low firmness). The prediction is done in two steps. First, a full range model (full model) is applied. When the full model prediction gives a low firmness value, then the over-ripe model is applied to refine the prediction. The full model is reasonably significant in regression terms, robust, but allows only a coarse quantitative prediction (standard deviation ratio, SDR = 2.48, 1.50 and 2.40 for sets 1, 2 and 3, respectively). Also, RMSEP% = 139%, 91% and 56%, indicating large relative errors at low firmness values. The segmented model improved moderately the correlation, and the values of RMSEC, RMSEP and SDR; it improved significantly the RMSEP% (29%, 55% and 31%), providing an improvement of the relative prediction errors at low firmness values. This method improves the ordinary PLS models. Finally, we tested whether chlorophyll alone was enough for a predictive model for firmness, but the results showed that the absorption of chlorophyll alone does not explain the performance of the PLS models.     

Hyperspectral laser-induced fluorescence imaging for assessing apple fruit quality

Chlorophyll fluorescence is a promising technique for postharvest quality research, and reported studies were mainly based on the fluorescence kinetic analysis method, which has not been quite satisfactory for measuring fruit quality. This paper reports on using a hyperspectral imaging technique for measurement of laser-induced fluorescence from apple fruit for predicting multiple fruit quality parameters. A continuous wave blue laser (408 nm) was used as an excitation source for inducing fluorescence in apples. Fluorescence scattering images were acquired from ‘Golden Delicious’ apples by a hyperspectral imaging system at the instance of laser illumination (0 min) and after 1, 2, 3, 4, and 5 min of illumination. The hyperspectral fluorescence image data were represented by mean, maximum, and standard deviation spectra. Standard tests were performed on measuring fruit skin and flesh color, firmness, soluble solids content, and titratable acid. A hybrid method of combining principal component analysis and neural network modeling was used for developing prediction models to predict fruit quality parameters for each of the six illumination time periods. Fluorescence emission decreased steadily during the first 3 min of illumination and was stable within 5 min. The 0-min fluorescence prediction models had somewhat poorer prediction results for individual quality parameters except skin hue than did the models of longer illumination time. The differences were minimal in the model prediction results from the fluorescence data at 1, 2, 3, 4 or 5 min of illumination. Overall, excellent predictions were obtained for apple skin hue with the correlation coefficient of prediction of 0.94. Relatively good predictions were obtained for fruit firmness, skin chroma, and flesh hue with values for the correlation coefficient being equal to or greater than 0.74 for 1 min of illumination or longer times, and poorer correlations were found for soluble solids content, titratable acid, and flesh chroma. Hyperspectral laser-induced fluorescence imaging is potentially useful for assessing selected quality parameters of apple fruit and further research is needed to improve fluorescence measurement for better prediction of fruit quality.     

Factors that affect tomato bruise development as a result of mechanical impact

External and internal factors that affect tomato bruise susceptibility such as impact- and fruit-related properties were investigated. Logistic regression was used to establish a relationship between tomato loading conditions and the resulting damage. Impact-specific models were built for a more precise determination of the bruise risk related to a narrow range of impact energies, being low (23 mJ), medium (71 mJ), high (158 mJ) and very high (216 mJ) impacts.Pericarp tissue over the locules was much more sensitive to bruise development than radial wall tissue. Tomatoes at room temperature (20 °C) were more sensitive than fruit stored at 12 °C. Tomato susceptibility to bruising increased substantially with ripening and loading conditions. The duration of the impact played a critical role in the bruise development and it is largely determined by fruit intrinsic parameters. Additional effects of the restitution coefficient and the fruit mass were found. Finally, different factors are responsible for tomato bruising in the various impact classes. The effect of low and medium energy impacts is largely controlled by the fruit texture. Especially medium impacts seem to substantially increase the bruise potential. The bruise potential of high and very high impacts mainly depends on fruit ripeness and the location of impact.     

SPME-GCMS integrated with chemometrics as a rapid non-destructive method for predicting microbial quality of minimally processed jackfruit (Artocarpus heterophyllus) bulbs

SPME-GCMS in combination with chemometrics was employed to correlate volatile headspace composition with microbial quality of minimally processed jackfruit (Artocarpus heterophyllus) bulbs stored at 4 °C and 10 °C. Predictive models of the total viable count (TVC) and yeast and mold count (Y&M) were prepared by Partial Least Square Regression (PLS-R) using total ion current (TIC) and total mass spectral data as independent variables. All PLS-R models correlating microbial quality with GC spectral data and total mass spectral data demonstrated high regression coefficient (R > 0.93). Models generated using TIC performed better in comparison with models prepared with total mass spectral data against test data. Ethanol, ethyl acetate and 3-methyl-1-butanol were identified as major compounds responsible for the observed correlations. The possibility of using GCMS as a nondestructive method for rapid assessment of microbial quality of minimally processed fruits is demonstrated here for the first time.     

Development of multispectral imaging algorithm for detection of frass on mature red tomatoes

In this research, a multispectral fluorescence-based imaging algorithm was developed to detect frass contamination on mature Campari tomatoes. Tomato images were acquired using a hyperspectral fluorescence line-scan imaging system with violet LED excitation, then analyzed for wavelength selection. The fluorescence intensities at five wavelengths, 515 nm, 640 nm, 664 nm, 690 nm, and 724 nm, were used to compute three simple ratio functions to detect frass contamination. The contamination spots were created on the tomato surfaces using four low-concentration frass dilutions. The algorithms detected over 99% of the 0.2 kg/L and 0.1 kg/L frass contamination spots and successfully differentiated these spots from tomato skin surfaces, stem scars, and stems. However, differentiation of the 0.05 kg/L and 0.02 kg/L frass contamination spots was more difficult. Adjusting the algorithm to successfully detect 95% of the 0.05 kg/L spots also resulted in false-positive pixel detections occurring on 28% of the tomatoes. This study demonstrates that a simple multispectral fluorescence imaging algorithm based on violet LED excitation could be useful for rapid postharvest detection of frass contamination on tomatoes in processing lines.     

Postharvest quality of apple predicted by NIR-spectroscopy: Study of the effect of biological variability on spectra and model performance

The effect of cultivar, season, shelf-life and origin on the accuracy of near infrared (NIR) calibration models for the soluble solids content (SSC) and firmness of apple was studied based on a large spectral data set based on approximately 6000 apple fruit from different cultivars, origins, shelf-life exposure time and seasons. To interpret the variance in the spectra with respect to biological variability, functional analysis of variance (FANOVA) was used. From the FANOVA analysis it was concluded that the effects of cultivar, origin and shelf-life exposure time on the NIR spectra were all significant. The largest differences in the spectra were found around the water absorption peaks (970, 1170 and 1450 nm). External validations using independent data sets showed that the accuracy of the models increased considerably when more variability was included in the calibration data set. In general the RMSEP for predictions of the SSC were in the range 0.6–0.8 °Brix, while for Magness Taylor firmness it was 5.9–8.8 N, depending on the cultivar. It was shown that atypical data can lead to large validation errors. It is, therefore, important to collect a calibration data set which is sufficiently representative for future samples to be analyzed with the developed calibration models and to develop simple procedures for model adaptation during practical use.     

Cold quarantine responses of ‘Tarocco’ oranges to short hot water and thiabendazole postharvest dip treatments

This study investigated the effects of brief hot water and thiabendazole (TBZ) postharvest dip treatments on ultrastructural changes of fruit epicuticular wax (ECW), TBZ residues, decay development and quality traits of ‘Tarocco’ oranges [Citrus sinensis (L.) Osbek] subjected to cold quarantine, subsequent simulated transport and shelf-life. Commercially mature fruit were submerged in water at 20 °C (control fruit) or TBZ at 1000 mg/L and 20 °C for 60 s, or in hot water without or with TBZ at 300 mg/L and 53, 56, or 59 °C for 60, 30, and 15 s respectively. Following treatments, fruit were stored for 3 weeks at 1 °C (simulated quarantine conditions for fruit disinfestations against Mediterranean fruit fly, Medfly), followed by 4 days at 3 °C (simulated long distance transport), and finally kept at 20 °C for 3 days (shelf-life, SL). Scanning electron microscopy (SEM) analysis of ‘Tarocco’ orange surface showed that the typical wax platelets, lifting around edges of wax plates and areas free of epicuticular wax (ECW), that disappeared after hot water dips at 53–59 °C for 60–15 s, become visible again after storage for 21 days at 1 °C (quarantine conditions), and changes involving the appearance of rough ultrastructure, presence large curled plates, fissured wax crusts, and areas with ECW deficiencies, became much more pronounced after shelf-life. These occurrences were related to the transient effect of hot water treatment in decay control. Conversely, treatments with 300 mg/L TBZ 53 °C for 60 s or 56 °C for 30 s effectively reduced decay after quarantine. These treatments were as effective as standard treatment with 1000 mg/L TBZ at 20 °C and produced similar TBZ residue levels in fruit, without impairing fruit quality traits such as visual appearance, weight loss, compression test, sensory attributes, juice color parameters (a*, b*, h, L*, and Chroma), and juice chemical characteristics (soluble solids content, titratable acidity, ascorbic acid, glucose, sucrose, citric acid, total phenols, total anthocyanins, and total antioxidant activity).     

Studies on classification models to discriminate ‘Braeburn’ apples affected by internal browning using the optical properties measured by time-resolved reflectance spectroscopy

This work aimed at studying the feasibility of time-resolved reflectance spectroscopy (TRS) to nondestructively detect internal browning (IB) in ‘Braeburn’ apples through the development of classification models based on absorption (μ_a) and scattering (μ_s′) properties of the pulp.This research was carried out in two seasons: in 2009, apples were measured by TRS at 670 nm and in the 740–1040 nm spectral range on four equidistant points around the equator, whereas in 2010 apples were measured by TRS at 670 nm and at 780 nm on eight equidistant points.The values of the absorption coefficients measured in the 670–940 nm range increased with IB development. On the contrary, μ_s′780 was higher in healthy fruit than in IB ones. The μ_a780 also significantly increased with IB severity, showing high values when IB affected the pulp tissues compared to the core ones. Also μ_a670 changed with IB development, but it was not able to clearly discriminate healthy fruit from IB ones because its value was also affected by the chlorophyll content of the pulp.The absorption and scattering coefficients were used as explanatory variables in the linear discriminant analysis in order to classify each apple tissue as healthy or IB; then the models obtained were used for fruit classification. The best classification performance was obtained in 2010 using μ_a780 and μ_s′780 and considering the IB position within the fruit: 90% of healthy fruit and 71% of IB fruit were correctly classified. By using all the μ_a measured in the 670–1040 nm range plus the μ_s′780, IB fruit classification was slightly better while healthy fruit classification was worse. The better result of 2010 was due to the increased number of TRS measurement points that allowed better exploration of the fruit tissues. However, the asymmetric nature of this disorder makes detection difficult, especially when the disorder is localized in the inner part of the fruit (core) or when it occurs in spots. A different TRS set-up (position and distance of fibers, time resolution) should be studied in order to reach the deeper tissue within the fruit in order to improve browning detection.     

Effect of fruit moving speed on predicting soluble solids content of ‘Cuiguan’ pears (Pomaceae pyrifolia Nakai cv. Cuiguan) using PLS and LS-SVM regression

Visible (Vis)/near infrared (NIR) spectroscopy is an excellent technique for non-destructive fruit quality assessment. This research was focused on evaluating the use of Vis/NIR spectroscopy for measuring soluble solids content (SSC) of intact ‘Cuiguan’ pears (Pomaceae pyrifolia Nakai cv. Cuiguan) on-line. Also, the effect of fruit moving speed on SSC measurements was investigated. Diffuse transmission spectra were collected using a fiber spectrometer equipped with a 3648-element linear silicon CCD array detector in the wavelength range of 345–1040 nm, and all sample spectra were collected three times at different fruit moving speeds of 0.3 m s^?1, 0.5 m s^?1 and 0.7 m s^?1. Spectral pre-processing such as derivative, standard normal variate transformation (SNV) and multiplicative scatter correction (MSC) was used before calibration. Partial least squares (PLS) and least squares support vector machines (LS-SVM) were used to develop calibration models for SSC. The results show that fruit moving speed has few effects on spectra and model performance at a fruit moving speed of 0.3–0.7 m s^?1. At 0.5 m s^?1, the best model for SSC was PLS regression coupled with original spectra, its coefficient of determination (R²) and root mean square error of prediction (RMSEP) being 0.916% and 0.530%, respectively.     

Investigating the effects of simulated transport vibration on tomato tissue damage based on vis/NIR spectroscopy

Visible and near infrared (vis/NIR) spectroscopy combined with chemometrics were investigated to evaluate the effects of simulated transport vibration levels on damage of tomato fruit. A total of 280 tomato samples were randomly divided into 5 groups; each group was subjected to vibration at different acceleration levels. A total of 230 samples (46 from each group) were selected as a calibration set; whereas 50 samples (10 from each group) were selected as a prediction set. Raw spectra, differentiation (the first derivative) spectra, extended multiplicative scatter correction (EMSC) processed spectra and standard normal variant combined with detrending (SNV–DT) processed spectra were used for calibration models. SNV–DT processed spectra had the best performance using for partial least squares (PLS) analysis. The PLS analysis was implemented to calibrate models with different wavelength bands including visible, short-wave near infrared (SWNIR) and long-wave near infrared (LWNIR) regions. The best PLS model was obtained in the vis/NIR (600–1600 nm) region. Using a grid search technique and radial basis function (RBF) kernel, four least squares support vector machine (LS–SVM) models with different latent variables (7, 8, 9, and 10 LVs) were compared. The optimal model was obtained with 9 LVs and the correlation coefficient (r), root mean square error of prediction (RMSEP) and bias for the best prediction by LS–SVM were 0.984, 0.137 and 0.003, respectively. The results showed that vis/NIR spectroscopy could be applied as a reliable and rapid method for predicting the effect of vibration levels on tissue damage of tomato fruit.     

Using visible and near infrared diffuse transmittance technique to predict soluble solids content of watermelon in an on-line detection system

Sugar content is one of the most important factors determining the eating quality of watermelon fruit. In order to detect the fruit soluble solids content (SSC) on-line, this work develops a nondestructive on-line detection prototype system using visible and near-infrared (Vis/NIR) technology. For the acquisition of the diffuse transmittance spectrum of watermelon, the conveyor was set at a speed of 0.3 m/s and ten 150 W tungsten halogen lamps were used as the light source. The crucial model for SSC value prediction was optimized by chemometrics. Partial least squares regression (PLSR), stepwise multiple linear regressions (SMLR), Monte-Carlo uninformative variable elimination (MC-UVE) and genetic algorithms (GA) were applied to the spectra in the range of 687–920 nm. The data pre-processing methods were optimized to transmittance spectra with baseline offset correction (BOC), and the BOC-MC-UVE-SMLR calibration model was the best with a correlation coefficient (r_pre) of 0.70, root mean square error of prediction (RMSEP) of 0.33 °Brix for the prediction set. In on-line testing of 30 samples, the r_pre was 0.66 and RMSEP was 0.39 °Brix. The results showed that a nondestructive on-line SSC value determination prototype based on Vis/NIR technology was feasible.