TY - JOUR T1 - Computerised texture image analysis of biocrystallisation images A1 - Andersen, Jens-Otto JA - Elem. d. Naturw. JF - Elemente der Naturwissenschaft PY - 2006 VL - 85 SP - 111 EP - 111 DO - 10.18756/edn.85.111 SN - p-ISSN 0422-9630 LA - en N2 -
The five main conclusions of my investigations were:
1) So far several studies using computerised image analysis have been published, which were able to discriminate in a relevant manner between specific samples.
2) The approaches can be divided into textural and more structural approaches.
3) In some studies results were improved when introducing more structural elements into a textural approach.
4) Due to the structural character of a biocrystallisation picture we may expect that ultimately a structural approach will be most succesful.
5) We may not expect that image analysis will be able to reflect the ‘Gestalt’ level of the samples, and is only a helpful tool in combination with visual evaluation.
The five main conclusions of my investigations were:
1) So far several studies using computerised image analysis have been published, which were able to discriminate in a relevant manner between specific samples.
2) The approaches can be divided into textural and more structural approaches.
3) In some studies results were improved when introducing more structural elements into a textural approach.
4) Due to the structural character of a biocrystallisation picture we may expect that ultimately a structural approach will be most succesful.
5) We may not expect that image analysis will be able to reflect the ‘Gestalt’ level of the samples, and is only a helpful tool in combination with visual evaluation.
The five main conclusions of my investigations were:
1) So far several studies using computerised image analysis have been published, which were able to discriminate in a relevant manner between specific samples.
2) The approaches can be divided into textural and more structural approaches.
3) In some studies results were improved when introducing more structural elements into a textural approach.
4) Due to the structural character of a biocrystallisation picture we may expect that ultimately a structural approach will be most succesful.
5) We may not expect that image analysis will be able to reflect the ‘Gestalt’ level of the samples, and is only a helpful tool in combination with visual evaluation.