Computer recognition of three-dimensional objects from optical images
Jeram Godhumal Advani · OhioLink ETD Center (Ohio Library and Information Network) · 1971
The dissertation has been concerned with the problem of classification of a solid object and the estimation of the six parameters characterizing its position from a discrete set of data points associated with the silhouette of a solid object. This has been achieved by the development of a computer oriented algorithm which can successfully classify and estimate the six parameters of a parallelepiped, a winged parallelepiped and an airplane. The feasibility of the recognition algorithm has been investigated by using the simulated data points of these three objects. The recognition scheme is finally applied to estimate the parameters of an aircraft from the data points generated on the boundary of a real picture. (Author)