METHOD AND DEVICE FOR THE CHARACTERIZATION AND LOCALIZATION IN REAL TIME OF SINGULAR FEATURES IN A DIGITALIZED IMAGE

摘要

The singularities constituting the characteristic points of the image contour points of high curvature, such as that of the corners or regions of small size, the image being formed of a two-dimensional array of pixels each having a gray level determined. (br / the device of the invention comprises: means derivative (10) approximating, for each pixel to be analyzed, the partial derivatives of order 2 of the function of the gray level of the image (f (x, y)); means caractérisateurs (23, 30), determining from these derivatives, an attribute (lambda2) representative of the sought-after characteristic and affecting this attribute to the pixel and analyzed, and of the localization means, searching among the pixels of the image thus analysed the points maximizing the attribute. & br / very advantageously, the differentiating means (10) comprise means convolvers performing convolution of the function of the gray level of the image (f (x, y)) with the functions corresponding to the partial derivative of order 2 of a function of regularization, this function of the controller making it possible to attenuate the noise present in the image. & br / preferably, the attribute chosen is the second value of the matrix of the derivatives seconds, and the function of regularization is a gaussian function, in particular a gaussian function centered.