Properties of multiscale morphological filters, namely, the morphology decomposition theorem

Chardaire, Pierre; Bangham, Jenny; Pye, C. Jeremy; Wu, Dequan

doi:10.1117/12.172552

Cited by 3 publications

(2 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Sieves are described in more detail elsewhere in this conference [12,13,14], but in short, a root median filter comprises a series of root median filters each of larger scale than the previous one. So, for example, a window of 3 root median filter feeds a window 5 root median filter and so forth.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Median and morphological scale space filtering and zero-crossings

Bangham

1994

SPIE Proceedings

Self Cite

View full text Add to dashboard Cite

Until recently, attention has been focused on linear methods for achieving multiscale decomposition, indeed certain filters have been considered uniquely appropriate in part because they, Gaussian filters for example, localise features well and partly because they do not create new inflexions with increasing scale. Unfortunately even filters, such as Gaussians, produce decompositions in which information, associated with edges and impulses, is spread over many, or all, scale space channels and this both compromises edge location and potentially pattern recognition.An alternative is to use nonlinear filter sequences (filters in series, known as sieves) or banks (in parallel). Recently multiscale decomposition using both erosion (dilation) and closing (opening) operations with sets of increasing scale flat structuring elements have been used to analyse edges over multiple scales and the granularity of images. These do not introduce new edges as scale increases. However, they are not at all statistically robust in the face of, for example, salt and pepper noise. This paper shows that sieves also do not introduce new edges, are very robust and perform at least as well as discreet Gaussian filters when applied to sampled data. Analytical support for these observations is provided by the morphology decomposition theorem discussed elsewhere in this volume.

show abstract

Section: Resultsmentioning

confidence: 99%

“…4 shows an example of pattern recognition in 1D at multiple scales. The signature of pattern A (trace I) is used to configure the pass band, in the granularity domain [13,14] and the result of passing the signal (trace II) through such a matched sieve is shown in Fig. 41.…”

Section: Resultsmentioning

confidence: 99%