Correlation analysis of hyperspectral imagery for multispectral wavelength selection for detection of defects on apples

Visible/near-infrared reflectance spectra extracted from hyperspectral images of apples were used to determine wavelength pairs that can be used to distinguish defect regions from normal regions on the apple surface. The optimal wavelengths were selected based on correlation analysis between the wavelength band ratio (λ1/λ2) or difference (λ1 − λ2) and the assigned value for the surface condition (0 = normal, 1 = defect). Spectral images of apple surfaces at the selected wavelengths were used to validate the correlation analysis. The correlation coefficients obtained using the correlation analysis for band ratio and difference were 0.91 and 0.79, respectively. When applied to the set of apple images, the band ratio model correctly identified 195 of the 211 defects on a set of 70 Fuji apples containing at least one defect region. Thus, the correlation analysis was demonstrated to be a feasible method for selecting wavelength pairs for use in distinguishing defects from areas without defects on apples.

[1]  Y. R. Chen,et al.  HYPERSPECTRAL REFLECTANCE AND FLUORESCENCE IMAGING SYSTEM FOR FOOD QUALITY AND SAFETY , 2001 .

[2]  M.L. Huebschman,et al.  Characteristics and capabilities of the hyperspectral imaging microscope , 2002, IEEE Engineering in Medicine and Biology Magazine.

[3]  W. R. Windham,et al.  Hyperspectral Imaging for Detecting Fecal and Ingesta Contaminants on Poultry Carcasses , 2002 .

[4]  Christopher W. Brown,et al.  Mechanisms for the conversion of hyperspectral images to chemical images , 2000, BiOS.

[5]  W S Grundfest,et al.  Consequences of scattering for spectral imaging of turbid biologic tissue. , 2000, Journal of biomedical optics.

[6]  Richard O. Claus,et al.  Cellular automata for the analysis of biomedical hyperspectral images , 2001, SPIE BiOS.

[7]  Charles R. Bostater.  Imaging derivative spectroscopy for vegetation dysfunction assessments , 1998, Remote Sensing.

[8]  N. Otsu.  A threshold selection method from gray level histograms , 1979 .

[9]  Daniel L. Farkas,et al.  Spectral imaging in biomedicine: a selective overview , 1998, Optics & Photonics.

[10]  Kurt C. Lawrence,et al.  Hyperspectral imaging for food processing automation , 2002, SPIE Optics + Photonics.

[11]  Charles A. DiMarzio,et al.  Wavelength selection for imaging hemoglobin in skin , 2000, SPIE Optics + Photonics.

[12]  K. Zuzak,et al.  Visible reflectance hyperspectral imaging: characterization of a noninvasive, in vivo system for determining tissue perfusion. , 2002, Analytical chemistry.

[13]  Jerry A. Burman,et al.  Hybrid pattern recognition method using evolutionary computing techniques applied to the exploitation of hyperspectral imagery and medical spectral data , 1999, Remote Sensing.

[14]  Bosoon Park,et al.  COLOR IMAGE CLASSIFICATION SYSTEMS FOR POULTRY VISCERA INSPECTION , 1999 .

[15]  Yoshio Inoue,et al.  Hyperspectral reflectance measurements for estimating eco-physiological status of plants , 2001, SPIE Asia-Pacific Remote Sensing.

[16]  Charles A. DiMarzio,et al.  Hyperspectral imaging for dermal hemoglobin spectroscopy , 1999, Optics & Photonics.

[17]  Mark A. Folkman,et al.  Application of hyperspectral imaging spectrometer systems to industrial inspection , 1996, Other Conferences.

[18]  Y. R. Chen,et al.  Detection of Defects on Selected Apple Cultivars Using Hyperspectral and Multispectral Image Analysis , 2002 .

[19]  Junichi Sugiyama,et al.  Near-infrared imaging spectroscopy based on sugar absorption band for melons. , 2002, Journal of agricultural and food chemistry.

[20]  Patrick Wambacq,et al.  Hyperspectral image sensor for weed-selective spraying , 1999, International Symposium on Photonics and Applications.