Abstract | ||
---|---|---|
A morphology-based thresholding technique for char- acter extraction from outdoor images is developed. Such extraction is traditionally difficult because of the influ- ence of illumination changes, such as lightness, shadow, and reflection. Changes in character font, apparent size, and inclination also make extraction more difficult. Though there are some character reading systems for format and size determined characters, such as the auto- matic license plate reader, their strong assumptions limit their applications. We assume for character extraction merely that characters occupy plane areas and consist of narrow lines. The proposed method makes it possible to choose an appropriate threshold for each point using the gray level of the nearest edge points. The method yields more robust thresholding results against changes in illumination conditions and character. |
Year | Venue | Keywords |
---|---|---|
1992 | MVA | mathematical morphology |
DocType | Citations | PageRank |
Conference | 0 | 0.34 |
References | Authors | |
1 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Yasuko Takahashi | 1 | 5 | 2.79 |
akio shio | 2 | 15 | 3.05 |
Kenichiro Ishii | 3 | 44 | 22.43 |