Adaptive Multi-scale Edge Detection in the EMD Domain

A multi-scale edge detection method in the empirical mode decomposition (EMD) domain using the consistent EMD method to decompose the reference noise to adaptively estimate the noise threshold is proposed to detect the edges of noisy signals with different signal to noise ratios. The new algorithm first calculates the slope of the EMD residuals in scales. Second, it eliminates the noise in the slope signal by thresholding, and the noise threshold is estimated according to the mean square value of the reference noise. At last, the edge information is output by the way of the spatial consistence test. Simulation results show that, the adaptive multi-scale edge detection method in the EMD domain can accurately detect the signal edges, and suppress the noise.