UDC 004.932
DETECTOR OF GRADIENT TYPE BORDERS FOR UNDERSTANDING SURFACE IMAGES
A. I. Novikov, Ph.D., associate professor, RSREU, Ryazan, Russia; This email address is being protected from spambots. You need JavaScript enabled to view it.
A. V. Pronkin, student of RSREU, Ryazan, Russia; This email address is being protected from spambots. You need JavaScript enabled to view it.
A new method of gradient type is proposed for highlighting the brightness differences in images. Estimation of partial derivatives in the composition of gradient vector in each pixel of the image is carried out using a line mask that provides smooth values of partial derivatives. The aim of the work is to create a new method for detecting the boundaries of brightness difference, which would have minimal computational complexity, stably select the boundaries in the conditions of moderate noise without using pre-filtering of random component, allow suppressing short «weak» lines in the images and form thin outlines of objects on this basis. The method was created as an alternative to the well-known Kenny method (Canny) for use in onboard vision systems of aircraft and robotic complexes. Its area of application may be much wider. The
proposed method is multi-step, which is due to the multiplicity of restrictions that it must satisfy.
Key words: image, partial derivative, gradient, gradient modulus, gradient vector direction, boundary detection, strong lines, weak lines, contour image.