Abstract | ||
---|---|---|
The large quantity of memory used for the process of producing the integral image, and when detecting the object using the integral image, has a lot to do with the large number of calculations that are performed. In this paper, we focus on computations when detecting the object using the integral image. In order to solve the problems typical of using the integral image, we propose the directional integral image, which gives a direction to the integral image. By using the directional integral image in the object detection procedure, we are able to perform calculations actually used in the object pattern region, and so the computational quantity of the directional integral image was decreased, compared with using the integral image. |
Year | DOI | Venue |
---|---|---|
2013 | 10.1109/URAI.2013.6677368 | URAI |
Keywords | Field | DocType |
integral image,object pattern region,directional integral image,object detection | Object detection,Computer vision,Feature detection (computer vision),Computer science,Artificial intelligence,Computation | Conference |
ISSN | ISBN | Citations |
2325-033X | 978-1-4799-1195-0 | 1 |
PageRank | References | Authors |
0.38 | 0 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Daeha Lee | 1 | 8 | 3.69 |
Jaehong Kim | 2 | 194 | 22.43 |