Visual detection of apple bruises using AdaBoost algorithm and hyperspectral imaging

Abstract

Hyperspectral imaging technique (400-1000 nm) was used for rapid and nondestructive recognition of bruises of apples. A total of 324 hyperspectral images were collected from 108 Fuji apples and the average spectral reflectance was extracted from the region of interest (ROI) of each image. The classification results of AdaBoost for the data pretreated by various existing methods were compared. Then, the correlation-based feature selection (CFS) algorithm was used to obtain characteristic wavelengths for reducing data redundancy. After pretreating with multiplicative scatter correction (MSC) and CFS, the average accuracy of the selected wavelengths was 97.63%. Then, an image processing algorithm based on the characteristic wavelengths selected before was proposed for the visual discrimination of bruises. This algorithm performed independent component analysis (ICA) transformation of the selected wavelengths, and chose the third component image of the ICA transform, then used adaptive threshold segmentation to obtain the bruise region of apples. The results showed that hyperspectral imaging technology could discriminate apple bruise, and this study can help to develop an online apple bruises detection system. ARTICLE HISTORY

A novel hyperspectral line-scan imaging method for whole surfaces of round shaped agricultural products

“…[68]. The sugar content of apples was reflected via the absorption valley at 820 nm [37]. When an apple bruise develops, the cells in the apple tissue are damaged and the intercellular air spaces decrease, causing differences in water content between bruised and sound regions [69].…”

“…This study used a post-processing method that discriminated sound regions from bruised regions, based on the glare from the apple surface, with an accuracy of 98%, see Figure 1.2-5. Bruised regions on 'Golden Delicious' apples were detected by multiple wavebands between 400 and 1000 nm [35], and the Ada-Boost algorithm were developed for detecting bruises on Fuji apples through visible and near infrared wavelengths [37], see Figure 1.2-6. These studies focused on determining only one, or the best, key-wavelength for bruise detection.…”

Review: Application of Artificial Intelligence in Phenomics

“…Another challenge in hyperspectral data analysis is data redundancy due to the continuous nature of wavelengths and their similarity. This has been alleviated by the selection of effective wavelengths using algorithms such as the successive projections algorithm (SPA), genetic algorithm (GA), the Monte-Carlo uninformative variable elimination (MC-UVE), and boosted regression tree (BRT) which is also a ML technique [ 81 , 82 , 83 ].…”

Integration of Partial Least Squares Regression and Hyperspectral Data Processing for the Nondestructive Detection of the Scaling Rate of Carp (Cyprinus carpio)

“…While applying this technology, it is necessary to implement a preprocessing method to reduce the interference signal and highlight the characteristic information of the spectra. Savitzky-Golay (SG) [24], first derivative (FD) [25], multiple scattering correction (MSC) [26], and standard normal variate (SNV) [27] methods have been proven to be effective in spectrum pretreatment. It was shown that these methods could reduce the influence of external factors and improve the detection accuracy to some extent [28].…”