Skip to main content
ARS Home » Research » Publications at this Location » Publication #129775
Title: MULTISPECTRAL DETECTION OF FECAL CONTAMINATION ON APPLES BASED ON HYPERSPECTRAL IMAGERY - PART I: APPLICATION OF VISIBLE-NEAR INFRARED REFLECTANCE IMAGING

Author
item Kim, Moon
item Chao, Kuanglin - Kevin Chao
item Chen, Yud
item Chan, Diane
item KIM, INTAEK - ISL-VISITING SCIENTIST
item Lefcourt, Alan

Submitted to: Transactions of the ASAE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/1/2002
Publication Date: 12/1/2002
Citation: KIM, M.S., CHAO, K., CHEN, Y.R., CHAN, D.E., KIM, I., LEFCOURT, A.M. MULTISPECTRAL DETECTION OF FECAL CONTAMINATION ON APPLES BASED ON HYPERSPECTRAL IMAGERY - PART I: APPLICATION OF VISIBLE-NEAR INFRARED REFLECTANCE IMAGING. TRANSACTIONS OF THE AMERICAN SOCIETY OF AGRICULTURAL ENGINEERS. 2002.

Interpretive Summary: This investigation illustrated a systematic approach using hyperspectral reflectance imaging technique in 450 to 851 nm regions of the spectrum in conjunction with the use of principal component analysis to define several optimal wavelengths to detect fecal contaminated spots on apples. We identified three visible to near-infrared (NIR) wavelengths and, alternatively, two NIR wavelengths that could potentially be used by multispectral imaging systems to detect fecal contamination of apples. This was demonstrated in the presence of large variations (reflectance intensity) of normal apples due to inherent morphological and skin coloration factors. Results indicated that three wavelengths in the green, red, and NIR region and two NIR bands, respectively, can be used to discriminate the fecal contaminated spots from uncontaminated apples surfaces. These optimal bands can be implemented to rapid on-line imaging systems for detection of fecal contamination. We can classify fecal contamination spots on apples for individual apple cultivars by use of the single threshold method. However, the reflectance imaging method detected only a fraction of thin feces spots indicating the lack of sensitivity. Further research is needed for the development and evaluation of algorithms (using selected wavelengths) that can be incorporated with multispectral imaging techniques for real-time classification of fecal contamination on apples. This research is useful to government and industry scientists and engineers who are developing noninvasive sensor systems for detection of contaminations on food commodities. Furthermore, multispectral bands and imaging techniques can be implemented in food processing plants to provide a rapid means of detecting fecal contamination on food produce.

Technical Abstract: Fecal contamination of apples is an important food safety issue. To allow development of automated methods to detect such contamination, a recently developed hyperspectral imaging system with a range of 450 to 850 nm was used to examine reflectance images of experimentally contaminated apples. Fresh feces from dairy cows was applied simultaneously as a thick patch and das a thin, transparent (not readily visible to the human eye), smear to four cultivars of apples, 'Red Delicious', 'Gala', 'Fuji' and 'Golden Delicious.' To address differences in coloration due to environmental growth conditions, apples were selected to represent the range of green to red colorations. Hyperspectral images of the apples and fecal contamination sites were evaluated using principal component analysis with the goal of identifying two-to-four wavelengths that could potentially be used in an on-line multispectral imaging system. Results indicate that contamination could be identified using either three wavelengths in the green, red, and NIR regions, or two wavelengths at the extremes of the NIR region under investigation. The three wavelengths in the visible to near infrared offer the advantage that the acquired images could also be used commercially for color sorting. However, detection using the two NIR wavelengths was found to be less sensitive to variations in apple coloration. For both sets of wavelengths, thick contamination could easily be detected using a simple threshold unique to each cultivar. In contrast, results suggest that more computationally complex analyses, such as combining threshold detection with morphological filtering, would be necessary to detect thin contamination spots using reflectance imaging techniques.