Metric Perimeter Calculation for Binary Convex Images Based on Shape and Edge Features
Yang Ren-er · Journal of Ningbo University · 2008
The paper presents an approach for computing the perimeter of the binary convex images by extracting features of shape and edge in metric unit.Firstly,the conversion is taken for calculating the ratio of mm/pixel.Several geometrical parameters concerning shape and contour are sought on the given binary image in order to construct a multiple linear regression model.The obtained model is designed to first estimate the perimeter displacement of the convex images,and eventually the desired perimeter.The test results indicate that the model achieves adequately high precision for the given samples,and is feasible for implementation.It is expected that this work can serve as an alternative approach for the related image analysis.