Image Features From Phase CongruencyPeter

Image features such as step edges, lines and Mach bands all give rise to points where the Fourier components of the image are maximally in phase. The use of phase congruency for marking features has signiicant advantages over gradient based methods. It is a dimension-less quantity that is invariant to changes in image brightness or contrast, hence it provides an absolute measure of the signiicance of feature points. This allows the use of universal threshold values that can be applied over wide classes of images. This paper presents a new way of calculating phase congruency through the use of wavelets. The existing theory that has been developed for 1D signals is extended to allow the calculation of phase congruency in 2D images. It is shown that for good localization it is important to consider the spread of frequencies present at a point of phase congruency. An eeective method for identifying, and compensating for, the level of noise in an image is presented. Finally, it is argued that high-pass ltering should be used to obtain image information at diierent scales. With this approach the choice of scale only aaects the relative signiicance of features without degrading their localization. Abstract Image features such as step edges, lines and Mach bands all give rise to points where the Fourier components of the image are maximally in phase. The use of phase congruency for marking features has signiicant advantages over gradient based methods. It is a dimensionless quantity that is invariant to changes in image brightness or contrast, hence it provides an absolute measure of the signiicance of feature points. This allows the use of universal threshold values that can be applied over wide classes of images. This paper presents a new way of calculating phase congruency through the use of wavelets. The existing theory that has been developed for 1D signals is extended to allow the calculation of phase congruency in 2D images. It is shown that for good localization it is important to consider the spread of frequencies present at a point of phase congruency. An eeective method for identifying, and compensating for, the level of noise in an image is presented. Finally, it is argued that high-pass ltering should be used to obtain image information at diierent scales. With this approach the choice of scale only aaects the relative signiicance of features without degrading their localization.