## Errors in binary images and an L p version of the Hausdorff metric (1992)

Citations: | 14 - 2 self |

### BibTeX

@MISC{Baddeley92errorsin,

author = {A.J. Baddeley},

title = {Errors in binary images and an L p version of the Hausdorff metric},

year = {1992}

}

### Years of Citing Articles

### OpenURL

### Abstract

This paper introduces a metric on the class of binary images, which can be used to assess the quality of image processing algorithms. Traditionally such assessment is done using the `statistical' misclassification error rate, or Pratt's figure of merit. However, we show that these measures have practical weaknesses. Further we show that they have undesirable topological properties in the continuous case. The classical Hausdorff metric generates the desired topology (the myopic topology on compact subsets) but is highly sensitive to noise and thus unsuitable for image processing purposes. We introduce a metric \Delta which is an L