碩士 / 國立臺灣科技大學 / 設計研究所 / 95 / This study discusses the design of pictograms in transportation sites from the theory of pattern recognition, with the goal to understand the “graphical element”, “design categories”, and “pattern recognition model combinations” of pictogram designs in transportation sites through inductive analysis. The study took place based on an investigation including five different stages: stage 1: the investigation on types and functions of pictograms in transportation sites; stage 2: the investigation on the contents expressed by pictograms; stage 3: the establishment and investigation on pictogram design categories; stage 4: the investigation on the effect of pattern recognition theory on the recognition of pictogram designs; and stage 5: the investigation on recognition confusion of pictograms in transportation sites through the method of “confusion matrix”. The following were found from the results of the research:
1.The functions of the different pictograms in transportation sites are mostly information signs. A minimum of one unit and a maximum of nine units of graphical elements form a pictogram, amongst which pictograms formed by four or less graphical units, occupied 94.12% of the entire pictogram set, indicating that pictogram styles formed according to different design requirements are able to comply with the “concise” design principle to pass the message around.
2.The graphical elements of the pictograms collected from MRT Taipei Main Station, Taiwan Railway Taipei Main Station, and the National Freeway Coach System were sorted by means of content analysis method where five major design categories: “Articles”, “Situation”, “Function”, “Action”, and “Site” were developed. The combinations of design categories for identification signs are most often expressed by means of a single “site category”; direction signs are most often expressed by means of a single “article category”; information signs are mostly expressed by means of the combination of “Article + Action” categories; and warning signs are mostly expressed by means of “Situation + Function + Action” categories.
3.As for the interactive use between design categories and pattern recognition models: the recognition of “article categories” is often processed by means of “feature matching” model; the recognition of “situation categories” processed by means of “feature matching”; “function category” by means of “template matching”; “action category” by means of “prototype matching”; and “site category” by means of “template matching”. The majority of pattern recognition model combination for the recognition of information signs is generally the use of simple “template matching”; for the recognition of direction signs the combination of “prototype matching + feature matching” is most often used; the combination of "prototype matching + feature matching” is most often used for the recognition of information signs; and the combination of “template matching + prototype matching + feature matching” is most often used for the recognition of warning signs.
4.The degrees of recognition for “design category combination” and “pattern recognition model combination” in the pictograms with enjoying high user recognition rate that appear to have similar results in rank and combinations with the "design category combination” and “pattern recognition model combination” in the pictograms currently and commonly used in transportation sites.
This study discovers that the formation and combination of graphical elements in a pictogram could use design category combinations as references for the future design thoughts. The use of pattern recognition models in a pictogram could help build model categories in the research for the recognition of pictograms, and could be discussed in more details and used in the future pictogram researches, in an attempt to increase the efficiency of the study and design of pictograms.
| Identifer | oai:union.ndltd.org:TW/095NTUS5619034 |
| Date | January 2007 |
| Creators | Chiang-hui Lin, 林江輝 |
| Contributors | none, 王韋堯 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 186 |
Page generated in 0.0114 seconds