Quantcast

Image processing technique to estimate geometric parameters and volume of selected dry beans

Research paper by Mahesh Kumar, Ganesh Bora, Dongqing Lin

Indexed on: 28 May '13Published on: 28 May '13Published in: Journal of Food Measurement and Characterization



Abstract

The geometric parameters along with related physical properties of agricultural material are vital to characterize and describe its quality. The application of image processing technique for this purpose can certainly reduce the human drudgery while ensuring quality of produce. The experiments were conducted to classify the shape and then workout the volume of the selected beans using the image processing technique. Green pea, garbanzo, kidney bean, navy bean and pinto bean were procured from the local grocery store for this study. A digital camera was used to capture the images of the ten different beans of each type. The beans were placed in two different orientations (longitudinal and lateral) and image processing technique was used to quantify and process the digital images. The results of image analysis were compared with the data obtained with actual measurements using digital vernier caliper. The related observations like thousand grain weight, bulk density, true density, porosity and weight of selected single grain were also recorded. A linear relationship was seen between the axial dimensions of beans and the pixel value with R2 in the range of 0.64–0.96. The bulk density and true density of the beans were observed to be in the range of 0.71–0.80 and 1.21–1.29 g/cc respectively. The sphericity of the beans varied in the range of 0.55–0.89. Analyses of the acquired images indicate convex geometry for the beans with ellipsoid shape while the same observation was recorded by physical measurements also. A linear relationship was observed between the volumes estimated by image analysis and true volumes of the beans with R2 in the range of 0.80–0.96. The circularity and compactness of the beans lied in the range of 0.69–0.90 and 0.56–0.71.