Image thinning with a cellular neural network
Takashi Matsumoto, Leon Ong Chua, T. Yokohama · IEEE Transactions on Circuits and Systems · 1990
The thinning problem requires that two tasks be implemented: (i) peeling the thick pixels off; (ii) stopping the peeling process when the pixel size reduces to exactly one. The first part can be achieved with relative ease. The main difficulty lies in the second part, because the stopping decision must be done automatically. It is shown that image thinning can be achieved in real-time using a cellular neural network (CNN) with eight planes, each one defined by a set of peeling templates and a set of stopping templates. The thinning processor, which is both analog and parallel, is described and examples are shown.>