Morphological filters of binary image and quantitative evaluation on filtering result

The characteristics of several morphological filtering models were analyzed and compared. Based on these models, a new morphological filtering model named as morphological thinning and thickening filter was presented. The set symmetric difference distance was made as an index to evaluate the filtering results quantitatively. The test result has proved that the new filter model is effective in removing linear noises both on background and in target, and it is superior to other filters in terms of quantitative evaluation index.