Early detection of bruises on apples is important for an automatic apple sorting system. A hyperspectral imaging system with the wavelength range of 400 to 1000nm was built for detecting bruises happened in an hour on ‘Fuji’ apples. Principal components analysis (PCA) was conducted on the hyperspecrtral images and the principal components (PC) images were compared. Three effective wavelengths 780, 850 and 960nm were determined using the weighing coefficients plot of the best PC image. Then, a multi-spectral imaging system with three bands 780, 850 and 960nm in the near-infrared range was developed. The system was consisted of two beamsplitters at 805 and 900nm, two bandpass filters and halogen tungsten lamp, and three CCD cameras. Images of 20 intact and 20 bruised apples were acquired. PCA was conducted on the three-band images of each apple and the best PC image was selected for bruise detection. A bruise detection algorithm based on the PC images and a global threshold method was developed. Results show that 90% of the bruised apples are correctly recognized.
Attribute of apple according to geographical origin is often recognized and appreciated by the consumers. It is usually an important factor to determine the price of a commercial product. Hyperspectral imaging technology and supervised pattern recognition was attempted to discriminate apple according to geographical origins in this work. Hyperspectral images of 207 Fuji apple samples were collected by hyperspectral camera (400-1000nm). Principal component analysis (PCA) was performed on hyperspectral imaging data to determine main efficient wavelength images, and then characteristic variables were extracted by texture analysis based on gray level co-occurrence matrix (GLCM) from dominant waveband image. All characteristic variables were obtained by fusing the data of images in efficient spectra. Support vector machine (SVM) was used to construct the classification model, and showed excellent performance in classification results. The total classification rate had the high classify accuracy of 92.75% in the training set and 89.86% in the prediction sets, respectively. The overall results demonstrated that the hyperspectral imaging technique coupled with SVM classifier can be efficiently utilized to discriminate Fuji apple according to geographical origins.
To effectively extract defective areas in fruits, the uneven intensity distribution that was produced by the lighting system or by part of the vision system in the image must be corrected. A methodology was used to convert non-uniform intensity distribution on spherical objects into a uniform intensity distribution. A basically plane image with the defective area having a lower gray level than this plane was obtained by using proposed algorithms. Then, the defective areas can be easily extracted by a global threshold value. The experimental results with a 94.0% classification rate based on 100 apple images showed that the proposed algorithm was simple and effective. This proposed method can be applied to other spherical fruits.
Early detection of bruises on apples is important for an automatic apple sorting system. A hyperspectral imaging system with the wavelength range of 1000 to 2500nm was built for detecting bruises happened in an hour on ‘Fuji’ apples. Principal components analysis (PCA) was conducted on the hyperspecrtral images and the principal components images were compared. Three effective wavelengths 1060, 1329 and 1949nm were determined using the weighing coefficients plot of the best principal component (PC) image. A bruise detection algorithm based on PCA on the three effective wavelengths and a global threshold method was developed. Independent validation set of 50 intact and 50 bruised apples was used to evaluate the performance of the developed algorithm. Results show that 100% of the intact apples are correctly classified, 94% of the bruised apples are correctly recognized and the overall detection accuracy is 97%.
KEYWORDS: Calibration, Solids, Near infrared, Near infrared spectroscopy, Nondestructive evaluation, Spectroscopy, Performance modeling, Data modeling, Solid modeling, Chemical analysis
This paper indicates the feasibility to use near infrared (NIR) spectroscopy combined with synergy interval partial least squares (siPLS) algorithms as a rapid nondestructive method to estimate the soluble solid content (SSC) in strawberry. Spectral preprocessing methods were optimized selected by cross-validation in the model calibration. Partial least squares (PLS) algorithm was conducted on the calibration of regression model. The performance of the final model was back-evaluated according to root mean square error of calibration (RMSEC) and correlation coefficient (R2c) in calibration set, and tested by mean square error of prediction (RMSEP) and correlation coefficient (R2p) in prediction set. The optimal siPLS model was obtained with after first derivation spectra preprocessing. The measurement results of best model were achieved as follow: RMSEC = 0.2259, R2c = 0.9590 in the calibration set; and RMSEP = 0.2892, R2p = 0.9390 in the prediction set. This work demonstrated that NIR spectroscopy and siPLS with efficient spectral preprocessing is a useful tool for nondestructively evaluation SSC in strawberry.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.