An image de-noising adaptive algorithm based on modified mathematical morphology operation

Two basic operations of the morphology,erosion and dilation,can be used to remove the image noise at the cost of severe information loss because of the inherent characteristics of these operations.Hence the morphology operation extracts the median instead of the traditional operations-minimum and maximum. Then the unit of structuring element is defined and the structuring element unit of zero matrices is deduced and found by mathematical method.Based on these work an adaptive image de-noising algorithm is presented.Compared with the median filter and the traditional mathematical morphology filter the new approach's performance is much better in the case of low signal-noise ratio.The new approach has wide application to the image de-noising field.