Location: Sugarbeet and Bean Research
Title: Evaluation of Internal Defect and Surface Color of Whole Pickles Using Hyperspectral Imaging Authors
|Ariana, Diwan -|
Submitted to: Journal of Food Engineering
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: September 3, 2009
Publication Date: September 9, 2009
Citation: Ariana, D.P., Lu, R. 2009. Evaluation of Internal Defect and Surface Color of Whole Pickles Using Hyperspectral Imaging. Journal of Food Engineering. 96(4):583-590. Interpretive Summary: Bloater damage often occurs to whole green-stock cucumbers during the brining process, which may have been caused by adverse production condition, nutrient deficiency during cucumber growth, and improper harvest and brining procedure. Bloated pickles typically exhibit soft or watery tissue and hollow center, and they are usually processed into low value final products. Currently, bloated pickles are separated from normal ones by human inspectors based on visual inspection and hand touch. This inspection method is labor intensive and unreliable due to speed demand and inspector fatigue. Therefore, development of an effective integrated inspection system that can detect both internal and external qualities of whole pickles would be valuable to the pickling industry. In this research, an on-line hyperspectral imaging prototype was used to separate defective whole pickles from normal, good ones. Hyperspectral imaging is a new generation imaging technology that enables obtaining both spectral and spatial information from the sample. The online imaging prototype was operated in simultaneous reflectance and transmittance modes. Reflectance model, covering the visible region of 400-675 nm, enabled to evaluate external characteristics of pickles (i.e., color, shape, size), whereas transmittance mode for the wavelengths of 675-1000 nm permitted light to pass whole pickles, thus enabling inspection of internal quality characteristics. Good and defective (bloater damage) whole pickles were collected from a commercial pickle processing facility. Hyperspectral images of these pickles were acquired using the online imaging prototype. Computer algorithms were developed to classify the pickles into two quality classes of 'normal' and 'defective'. The color of individual pickles was also evaluated from the reflectance signals. Results showed that the online imaging prototype in transmittance mode achieved an overall classification accuracy of 86%, compared with 70% accuracy obtained by the human inspectors. Color measurements by the online imaging system were not different from those measured by a standard colorimeter offline. The hyperspectral imaging system demonstrated superior capability of inspecting both internal and external quality characteristics of whole pickles. With further improvements, the technology should be useful for pickle quality inspection, thus helping the pickling industry in ensuring product quality and achieving labor cost savings.
Technical Abstract: Hyperspectral imaging operated under simultaneous reflectance (400-675 nm) and transmittance (675-1000 nm) 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 acquired using an on-line hyperspectral imaging prototype operating in the wavelength range of 400-1000 nm. Principal component analysis was applied to the hyperspectral images of 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 was 94.0 and 93.8 respectively. Transmittance images at 675-1000 nm were much more effective for internal defect detection compared to reflectance images for the visible region of 500-675 nm. 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.