Robust edge detection in noisy images using an adaptive stochastic gradient technique

The problem of edge detection in noisy images is addressed in this paper. It is shown that the performance of an existing edge detection method, known as the stochastic gradient operator, can be significantly improved by incorporating three new features: (i) a robust technique for estimating the noise variance and autocorrelation function of the signal, (ii) a block-by-block adaptation of the gradient mask, and (iii) calculation of a threshold based on Rayleigh distribution. The performance of the proposed technique is compared with that of some existing ones.