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. Hyperpectral 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. Published by Elsevier B.V.

Hyperspectral imaging is useful for detecting internal defects of pickling cucumbers. The technique, however, is not yet suitable for high-speed online implementation due to the challenges in analyzing large-scale hyperspectral images. This research aimed to select the optimal wavebands from the hyperspectral image data, so that they can be deployed in either a hyperspectral or multispectral imaging-based inspection system for the automatic detection of internal defects of pickling cucumbers. Hyperspectral reflectance (400–700 nm) and transmittance (700–1,000 nm) images were acquired, using an in-house developed hyperspectral imaging system running at two conveyor speeds of 85 and 165 mm/s, for 300 “Journey” pickling cucumbers before and after internal damage was induced by mechanical load. Minimum redundancy–maximum relevance (MRMR) was used for optimal wavebands selection, and the loadings of principal component analysis (PCA) were also applied for qualitatively identifying the important wavebands that are related to the specific features. Discriminant analysis with Mahalanobis distance classifier was performed for the two-class (i.e., normal and defective) and three-class (i.e., normal, slightly defective, and severely defective) classifications using the mean spectra and textural features (energy and variance) from the regions of interest in the spectral images at selected waveband ratios. The classification results based on MRMR wavebands selection were generally better than those from PCA-based classifications. The two-band ratio of 887/837 nm from MRMR gave the best overall classification results, with the accuracy of 95.1 and 94.2 % at the conveyor speeds of 85 and 165 mm/s, respectively, for the two-class classification. The highest classification accuracies for the three-class classification based on the optimal two-band ratio of 887/837 nm were 82.8 and 81.3 % at the conveyor speeds of 85 and 165 mm/s, respectively. The mean spectra-based classification achieved better results than the textural feature-based classification, except in the three-class classification for the higher conveyor speed. The overall classification accuracies for all selected waveband ratios at the low conveyor speed were slightly higher than those at the higher conveyor speed, since the low speed resulted in more scan lines, thus higher spatial resolution hyperspectral images. The identified two-band ratio of 887/837 nm in transmittance mode could be applied for fast real-time internal defect detection of pickling cucumbers.

The objective of this research was to measure the spectral absorption (mu(o)) and reduced scattering coefficients (mu(s)') of peaches, using a hyperspectral imaging-based spatially resolved method, for maturity/quality assessment. A newly developed optical property measuring instrument was used for acquiring hyperspectral reflectance images of 500 'Red-star' peaches. The mu(a) and mu(s)' spectra for 515 to 1,000 nm were extracted from the spatially resolved reflectance profiles using a diffusion model coupled with an inverse algorithm. The absorption spectra of peach fruit were marked with absorption peaks around 525 nm for anthocyanin, 620 nm for chlorophyll-b, 675 nm for chlorophyll-a, and 970 nm for water, while mu(s)' decreased monotonically with the increasing wavelength for most of the tested samples. Both mu(a) and mu(s)' were correlated with peach firmness, soluble solids content (SSC), and skin and flesh color parameters. Better correlation results for partial least squares models were obtained using the combined values of mu(a) and mu(s)' (i.e., mu(a) x mu(s)' and mu(eff)) than using mu(a) or mu(s)', where mu(eff) is the effective attenuation coefficient: mu(eff) = [3 mu(a)(mu(a) + mu(s)')](1/2). The results were further improved using least squares support vector machine models, with values of the best correlation coefficient for firmness, SSC, skin lightness, and flesh lightness being 0.749 (standard error of prediction or SEP = 17.39 N), 0.504 (SEP = 0.92 degrees Brix), 0.898 (SEP = 3.45), and 0.741 (SEP = 3.27), respectively. These results compared favorably to acoustic and impact firmness measurements, whose correlations with destructive measurements were 0.639 and 0.631, respectively The hyperspectral imaging-based spatially resolved technique is useful for measuring the optical properties of peach fruit, and it has good potential for assessing fruit maturity/quality attributes.

The objective of this research was to measure the absorption (mu(a)) and reduced scattering coefficients (mu(s)') of peaches, using a hyperspectral imaging-based spatially-resolved method, for their maturity/quality assessment. A newly developed optical property measuring instrument was used for acquiring hyperspectral reflectance images of 500 'Redstar' peaches. mu(a) and mu(s)' spectra for 515-1,000 nm were extracted from the spatially-resolved reflectance profiles using a diffusion model coupled with an inverse algorithm. The absorption spectra of peach fruit presented several absorption peaks around 525 nm for anthocyanin, 620 nm for chlorophyll-b, 675 nm for chlorophyll-a, and 970 nm for water, while mu(s)' decreased consistently with the increase of wavelength for most of the tested samples. Both mu(a) and mu(s)' were correlated with peach firmness, soluble solids content (SSC), and skin and flesh color parameters. Better prediction results for partial least squares models were obtained using the combined values of mu a and mu s' (i.e., mu(a) x mu(s)' and mu(eff)) than using mu(a) or mu(s)', where mu(eff) = [3 mu(a) (mu(a)+ mu(s)')](1/2) is the effective attenuation coefficient. The results were further improved using least squares support vector machine models with values of the best correlation coefficient for firmness, SSC, skin lightness and flesh lightness being 0.749 (standard error of prediction or SEP = 17.39 N), 0.504 (SEP = 0.92 (circle)Brix), 0.898 (SEP = 3.45), and 0.741 (SEP = 3.27), respectively. These results compared favorably to acoustic and impact firmness measurements with the correlation coefficient of 0.639 and 0.631, respectively. Hyperspectral imaging-based spatially-resolved technique is useful for measuring the optical properties of peach fruit, and it also has good potential for assessing fruit maturity/quality attributes.

Spectral scattering is useful for assessing the firmness and soluble solids content (SSC) of apples. In previous research, mean reflectance extracted from the hyperspectral scattering profiles was used for this purpose since the method is simple and fast, and also gives relatively good predictions. The objective of this study was to improve firmness and SSC prediction for 'Golden Delicious' (GD), 'Jonagold' (JG), and 'Delicious' (RD) apples by integration of critical spectral and image features extracted from the hyperspectral scattering images over the wavelength region of 500-1000 nm, using spectral scattering profile and image analysis techniques. Scattering profile analysis was based on mean reflectance method and discrete and continuous wavelet transform decomposition, while image analysis included textural features based on first order statistics, Fourier analysis, co-occurrence matrix and variogram analysis, as well as multi-resolution image features obtained from discrete and continuous wavelet analysis. A total of 294 parameters were extracted by these methods from each apple, which were then selected and combined for predicting fruit firmness and SSC using partial least squares (PLS) method. Prediction models integrating spectral scattering and image characteristics significantly improved firmness and SSC prediction results compared with the mean reflectance method when used alone. The standard errors of prediction (SEP) for GD, JG, and RD apples were reduced by 6.6, 16.1, 13.7% for firmness (R-pred-values of 0.87, 0.95, and 0.84 and the SEPs of 5.9, 7.1, and 8.7 N), and by 11.2, 2.8, and 3.0% for SSC (Rpred-values of 0.88, 0.78, and 0.66 and the SEPs of 0.7, 0.7, and 0.9%), respectively. Hence, integration of spectral and image analysis methods provides an effective means for improving hyperspectral scattering prediction of fruit internal quality. Published by Elsevier B.V..

An effective optical inspection system for detecting defective pickling cucumbers is needed to help the pickle industry deliver consistent pickled products to the consumer. This research was intended to measure the optical absorption and scattering properties of normal and internally defective pickling cucumbers, using hyperspectral imaging-based spatially-resolved technique. Spatially-resolved hyperspectral scattering images were acquired from 50 freshly harvested ‘Journey’ pickling cucumbers in the summer of 2008 before they were subjected to rolling under mechanical load to induce internal damage. The damaged cucumbers were imaged 1 h and 1 day after the mechanical stress treatment. Spectra of the absorption and reduced scattering coefficients for pickling cucumbers were extracted from the spatially-resolved scattering profiles, using an inverse algorithm for a diffusion theory model, for the spectral range of 700–1000 nm. It was found that within 1 h after mechanical damage, changes in the absorption and reduced scattering coefficients for the cucumbers were minimal. One day after mechanical damage, the absorption coefficient for the cucumbers increased significantly (at 5–10% level) for the wavelengths of 700–920 nm, whereas the reduced scattering coefficient decreased significantly for the wavelengths of 700–1000 nm (at 10% level). Overall mechanical damage caused greater changes in absolute value to the scattering properties than to the absorption properties. This research suggests that effective defect detection can be achieved by enhancing scattering characteristics measurement in the optical evaluation of pickling cucumbers.

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 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 was used to detect fruit fly-infested pickling cucumbers. Hyperspectral reflectance (450-740 nm) and transmittance (740-1,000 nm) images were acquired simultaneously for 329 normal (infestation free) and fruit flyinfested 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 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. Overall, the hyperspectral imaging system performed better than manual inspection, which had an overall accuracy of 75% and decreased significantly for smaller size cucumbers. This research demonstrated that hyperspectral imaging is potentially useful for detecting fruit fly-infested pickling cucumbers.

Internal physical damage such as carpel separation or hollow centers in pickling cucumbers is difficult to detect by human inspectors, and the current inspection procedure would require cutting open the fruit, which prohibits evaluation of individual cucumbers. Our recent research has demonstrated that hyperspectral transmittance imaging technique can provide an effective means for internal defect detection in pickling cucumbers. However the technique is still expensive and cannot meet the speed requirement in commercial cucumber processing facilities. Therefore, an alternative technique using laser scattering imaging to inspect internal quality of cucumbers was investigated in this research. A diode laser with the center wavelength of 808 nm was used in the experiment. Fifty fresh pickling cucumber samples were subjected to mechanical load to simulate stress caused by mechanical harvesting and subsequent handling. Scattering images generated at 0, 30, 45, 60, and 80 degrees of laser incident angle relative to the optical axis under transmittance mode were acquired from the pickling cucumbers before and two hours after mechanical stress was applied. Image processing and analysis algorithms were developed and tested on the acquired images to distinguish defective from normal cucumbers. Detection accuracies decreased as the incident angle increased. The best detection accuracy of 96% was achieved with 0 degree incident angle. The laser scattering technique could provide a cost effective way for rapid detection of internal defect of pickling cucumbers.

Spectral scattering is useful for assessing the firmness and soluble solids content (SSC) of apples. In previous research, mean reflectance extracted from the hyperspectral scattering profiles was used for this purpose since the method is simple and fast and also gives relatively good predictions. The objective of this study was to improve firmness and SSC prediction for 'Golden Delicious' (GD), 'Jonagold' (JG), and 'Red Delicious' (RD) apples by integration of critical spectral and image features extracted from the hyperspectral scattering images over the wavelength region of 500-1,000 nm, using spectral scattering profile and image analysis techniques. Scattering profile analysis was based on mean reflectance method and discrete and continuous wavelet transform decomposition, while image analysis included textural features based on first order statistics, Fourier analysis, co-occurrence matrix and variogram analysis, as well as multiresolution image features obtained from discrete and continuous wavelet analysis. A total of 294 parameters were extracted by these methods from each apple, which were then selected and combined for predicting fruit firmness and SSC using partial least squares (PLS) method. Prediction models integrating spectral scattering and image characteristics have improved firmness and SSC prediction results compared with the mean reflectance method when used alone. The standard errors of prediction (SEP) for GD, JG, and RD apples were reduced by 6.6, 16.1, 13.7% for firmness (R-values of 0.87, 0.95, and 0.84 and the SEPs of 5.9, 7.1, and 8.7 N), and by 11.2, 2.8, and 3.0% for SSC (R-values of 0.88, 0.78, and 0.66 and the SEPs of 0.7, 0.7,and 0.9 Brix), respectively.

Knowledge of the spectral absorption and scattering properties of apple tissue, especially bruised tissue, can help us develop an effective inspection method for detecting bruises during postharvest sorting and grading. This research was aimed at determining the optical properties of normal and bruised apple tissue for the wavelength range of 500-1,000 nm and quantifying their changes with time after bruising. Values for the absorption and reduced scattering coefficients were determined for the normal or unbruised tissue of 'Golden Delicious' and 'Red Delicious' apples and then for the bruised tissue, for different time intervals after bruising, using a hyperspectral imaging-based spatially resolved technique. Bruising caused changes to the absorption coefficient, but no consistent pattern of changes was observed for 'Golden Delicious' and 'Red Delicious' apples after they were bruised. The reduced scattering coefficient for normal apples, however, was much higher than that for bruised apples; it decreased consistently with time after bruising. These results suggest that bruising has a greater impact on scattering than on absorption. Hence, an optical system that enhances scattering feature measurement would be better suited for bruise detection.

Internal defect in pickling cucumbers can cause bloater damage during brining, which lowers the quality of final pickled products and results in economic loss for the pickle industry. Hence it is important to have an effective optical inspection system for detection and segregation of defective pickling cucumbers. This research was intended to measure the spectral absorption and scattering properties of normal and internally defective pickling cucumbers and whole pickles, using hyperspectral imaging-based spatially-resolved technique. Spatially-resolved hyperspectral scattering images were acquired from 50 freshly harvested 'Journey' pickling cucumbers in the summer of 2008. The cucumbers were then subjected to rolling under mechanical load to induce internal damage. The damaged cucumbers were imaged again one hour and one day after the mechanical stress treatment. In addition, 20 whole pickles each of normal and defective (bloated) class were also measured by following the same procedure as that for pickling cucumbers. Spectra of the absorption and reduced scattering coefficients for pickling cucumbers and whole pickles were extracted from the spatially-resolved scattering profiles, using an inverse algorithm for a diffusion theory model, for the spectral range of 700-1,000 nm. It was found that within one hour after mechanical damage, changes in the absorption and reduced scattering coefficients for the cucumbers were minimal. One day after mechanical damage, the absorption coefficient for the cucumbers increased noticeably for the wavelengths of 700-920 nm, whereas the reduced scattering coefficient decreased more significantly for the wavelengths of 700-1,000 nm. Overall mechanical damage had greater impact on the scattering properties than on the absorption properties. After brining, pickles became translucent and scattering was greatly diminished. Thus the diffusion theory model was no longer valid for determining the optical properties of whole pickles. This research suggests that effective defect detection may be achieved by enhancing scattering features in the optical evaluation of cucumbers.

Hyperspectral imaging under transmittance mode has shown potential for detecting internal defect, however, the technique still cannot meet the online speed requirement because of the need to acquire and analyze a large amount of image data. This study was carried out to select important wavebands for further development of an online inspection system to detect internal defect in pickling cucumbers and whole pickles. Hyperspectral transmittance/reflectance images were acquired from normal and defective cucumbers and whole pickles using a prototype hyperspectral reflectance (400–740 nm)/transmittance (740–1000 nm) imaging system. Up to four-waveband subsets were determined by a branch and bound algorithm combined with the k-nearest neighbor classifier. Different waveband binning operations were also compared to determine the bandwidth requirement for each waveband combination. The highest classification accuracies of 94.7 and 82.9% were achieved using the optimal four-waveband sets of 745, 805, 965, and 985 nm at 20 nm spectral resolution for cucumbers and of 745, 765, 885, and 965 nm at 40 nm spectral resolution for whole pickles, respectively. The selected waveband sets will be useful for online quality detection of pickling cucumbers and pickles.

Hyperspectral imaging under transmittance mode has shown promising results for detecting internal defect in pickling cucumbers, however, the technique still cannot meet the online speed requirement because it needs to acquire and process a large amount of image data. This study was conducted on selecting important wavebands as a basis for developing an online imaging system to detect internal defect in pickling cucumbers. 'Journey' pickling cucumbers were subjected to mechanical stress to induce damage in the seed cavity. Hyperspectral transmittance/reflectance images were acquired from normal and defective cucumbers using a prototype hyperspectral reflectance (400-740 nm)/ transmittance (740-1,000 nm) imaging system. Optimal wavelengths were determined by correlation analysis on single, ratio, and difference of two pairs of wavelengths. A global image thresholding method was applied to the selected spectral images to identify defective cucumbers. Images at 740 nm were the best for single waveband classification with an overall accuracy of 87%. For ratios of two wavebands, 925 nm and 940 nm resulted in an overall classification accuracy of 85%, and for differences of two wavebands, images at 745 nm and 850 nm were the best with a classification accuracy of 91%. All the selected wavebands were in the near infrared region, which is more effective for internal defect detection compared to the visible region.

Understanding optical properties of apple tissue, especially bruised tissue, can help us develop an effective method for detecting bruises during sorting and grading. This research was conducted on determining the optical properties of bruised apple tissue over 500-1,000 nm and quantifying their changes with time. Spectral absorption and reduced scattering coefficients were determined from the normal, unbruised tissues of 'Golden Delicious' and 'Red Delicious' apples and then bruised tissues at different time intervals after bruising, using a hyperspectral imaging-based spatially resolved technique. Absorption for normal tissues was generally lower than that for the bruised tissues in the spectral region of less than 600 nm, while an opposite trend was observed in the spectral region of 800-1,000 nm. The reduced scattering coefficient for normal tissues was much higher than that for the bruised tissues; it decreased consistently over time. Bruising had greater impact on scattering than on absorption. Hence an optical system that can enhance scattering features would be better suited for bruise detection.

Hyperspectral imaging operated under simultaneous reflectance (400–675nm) and transmittance (675–1000nm) modes was studied for non-destructive and non-contact sensing of surface color and bloater damage in whole pickles. Good and defective pickles were collected from a commercial pickle processing plant. Hyperspectral images of these pickles were obtained using a prototype of on-line hyperspectral imaging system, operating in the wavelength range of 400–1000nm. Principal component analysis was applied to the hyperspectral images of the pickle samples for bloater damage detection. Color of the pickles was modeled using tristimulus values calculated based on the hyperspectral images. There were no differences in chroma and hue angle of good and defective pickles. The average chroma of good and defective pickles was 15.5 and 15.0, respectively, and the hue angle 94.0° and 93.8°, respectively. Transmittance images at 675–1000nm were much more effective for internal defect detection compared to reflectance images for the visible region of 500–675nm. An overall defect classification accuracy of 86% was achieved, compared with an accuracy of 70% by the human inspectors. With further improvement, the hyperspectral imaging system could meet the need of bloated pickles detection in a commercial plant setting.

This article reports on the development of a hyperspectral imaging prototype for online evaluation of external and internal quality of pickling cucumbers. The prototype consisted of a two-lane round belt conveyor, two illumination sources (one for reflectance and one for transmittance), and a hyperspectral imaging unit. It had a novel feature of simultaneous imaging under reflectance mode in the visible region (400–675 nm) and transmittance mode for the red and near-infrared region (Red-NIR) (675–1000 nm). Reflectance information from the visible region was intended for evaluating the external characteristics of cucumbers such as skin color, whereas transmittance information from Red-NIR was used for internal defect detection (i.e., hollow center). Additional features of the prototype included simultaneous acquisition of reflectance and transmittance from calibration references that were installed in the system, to provide real-time, continuous corrections of individual hyperspectral images from each sample. Methods and algorithms were developed of estimating cucumber fruit size and correcting the effect of fruit size on transmittance measurements. The system was calibrated and evaluated for detecting the color, size, and internal defect of pickling cucumbers.

This paper reports on the development of a hyperspectral imaging prototype for evaluation of external and internal quality of pickling cucumbers. The prototype consisted of a two-lane round belt conveyor, two illumination sources (one for reflectance and one for transmittance), and a hyperspectral imaging unit. It had a novel feature of simultaneous imaging under reflectance mode covering the visible region of 400-675 nm and transmittance mode for 675-1000 nm, coupled with real-time, continuous calibration of reflectance and transmittance images for each cucumber using reference standards installed on the conveyor. Reflectance information was used for evaluating the external characteristics of cucumbers (i.e., skin color), transmittance for internal defect detection (hollow center), and the combined reflectance and transmittance for predicting flesh firmness. The prototype was tested on ‘Journey’ pickling cucumbers harvested in 2006 and 2007 for predicting skin and flesh color, flesh firmness, and internal defect. Hyperspectral images were processed to extract mean spectra for individual cucumbers, and partial least squares analysis was performed to predict flesh firmness, skin and flesh color, and the presence of internal defect. The prototype performed relatively well in predicting skin color with the coefficient of determination of 0.76 and 0.75 for chroma and hue respectively; however, it had poor prediction of flesh color and firmness. Transmittance data in the spectral region of 675-1000 nm provided excellent detection of internal defect for the pickling cucumbers, with the detection accuracy greater than 90%. The hyperspectral imaging technique would be useful for online inspection of surface color and internal defect on picking cucumbers.

Internal quality is an important aspect in the quality control and assurance of pickled products. A rapid and nondestructive method for internal defect detection would be of value to the pickle industry. A hyperspectral transmittance imaging technique was developed to detect internal defect in the form of carpel suture separation or hollow cucumbers resulting from dropping and rolling under load. Hyperspectral transmittance line scan images were collected from 'Journey' and 'Vlaspik' cucumbers over the six-day period after they were subjected to mechanical stress. Partial least squares discriminant analysis (PLS-DA) was performed on mean and standard deviation spectra extracted from the hyperspectral transmittance images to classify cucumber samples into defective or normal classes. A spectral-based pixel classification method using Euclidean distance was applied to classify pixels along the spatial dimension of the image into normal or defective class. The transmittance spectra of defective cucumbers were similar in shape to, but higher in magnitude than, those of normal cucumbers. Transmittance values for both defective and normal cucumbers were higher in the near-infrared range of 700-1000 nm than those in the visible range (450-700 nm). Average classification accuracies of 90.2%, 98.7%, and 95.4% were achieved using PLS-DA, whereas accuracies of 89.1%, 94.6%, and 90.5% were achieved using the spectral-based pixel classification method for Journey, Vlaspik, and the pooled data, respectively. The hyperspectral transmittance imaging technique can be used for rapid detection of internal defect in pickling cucumbers.