Fresh Green Peas

2. MATERIALS AND METHODS
2.1 Experimental Set up

Fresh green pea provided from local farm near Shiraz city, situated in the south-west of Iran. Care was taken when selecting green peas to obtain only those which approximately resemble a spherical shape. Size was measured using machine vision technique with a camera. Several measurements of the dimensions of the spherical samples were made and only samples within a 5% tolerance of the average dimensions were used. The average diameter of green peas was measured as 11.27 ± 0.35 mm. The thickness of the beds was equal to a single, double and triple green pea diameter for different treatments. To equilibrate the moisture content, green peas were kept in a plastic container in a refrigerator at

Schematic of cloudy sky chamber

Each 36-sample was photographed three times by a digital CCD camera (Canon IXUS 960 IS, 12MPX), then the best images with high quality and resolution were chosen and used. Images were transferred to a computer for processing with MATLAB 2011 software. Using MATLAB commands for extracting shape features, original images were converted into the gray image firstly, and then into the binary (black and white) images (Fig4).

Fig 4. Converting images into gray and binary

Five features (surface area, perimeter, maximum diameter, minimum diameter and equal diameter) were extracted directly from the images and later the shrinkage and roundness were calculated and evaluated by the following equations. The changes in the volumes of green peas before and after each experiment were determined. In addition – roundness - the fraction of largest projected area of object in natural rest position (Ap) to area of smallest circumscribing circle (Ac), was calculated (Mohsenin,

It should be noticed that the data extracted from MATLAB2011 is expressed in pixel which is then converted into millimeter. Excel (2010) software took the data directly from MATLAB. In the next stage the data were statistically analyzed in SPSS16 software. For color assessment, the images were also analyzed in the RGB and L*a*b* color spaces in MATLAB2011 software (Fig5). The word RGB is an abbreviation for red–green–blue colors. The RGB color space is based on RGB color model (Poynton and Charles, 2003). A particular RGB color space is based on the red, green, and blue, and can produce any color by these primary colors. The next used color space in this study was L*a*b* which is the most complete color space specified by the International Commission on Illumination. L*a*b* includes all the colors visible to the human eye whereas RGB includes about 90% all perceivable colors. L*a*b* is device independence to be used as a reference, This means that the colors are defined independent of their nature of creation or the device they are displayed on. In the L* a* b* color space, the component of L* represents the luminance of images which varies from 0 to 100 (L* = 0 yields black and L* = 100 indicates diffuse white), and components of a * is the rate of color change from green to red and b * the color