Object Recognition Supported by User Interaction for Service Robots
DOI: 10.1109/icpr.2002.1048379
|View full text |Cite
|
Sign up to set email alerts
|

Robust event detection by radial reach filter (RRF)

Abstract: We propose a novel srarisrical measrrre for robirsr event derecrion, called 'Radial Reuchfilrer' (RRF). The capabilin nfderecring new ohjecrs (evenrs)from a rime-series iniage is an imporranr problern of vision sysrems. The icsiral method of derecting nen objecrs is simple background subrrucrion rhar is ro subrrucr crrrrenr image frnm a backgroimd inrage. However: simple backgroicnd sirbrracrion is susceprible rr, illrrmiiinrion change such as shadows. And when rtie briglrrriess difference betweeti e w m and a… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
10
0

Publication Types

Select...
3
3
1

Relationship

0
7

Authors

Journals

citations
Cited by 21 publications
(10 citation statements)
references
References 5 publications
0
10
0
Order By: Relevance
“…(2) Real-time detection and refinement of human areas can be realized due to the high efficiency of the CENTRIST feature [4], and by detecting humans and computing an integral image only on the foreground extracted by the Radial Reach Filter (RRF) [6].…”
Section: Basic Ideasmentioning
confidence: 99%
See 2 more Smart Citations
“…(2) Real-time detection and refinement of human areas can be realized due to the high efficiency of the CENTRIST feature [4], and by detecting humans and computing an integral image only on the foreground extracted by the Radial Reach Filter (RRF) [6].…”
Section: Basic Ideasmentioning
confidence: 99%
“…1. First, RRF [6] is utilized to extract the foreground of each frame from an indoor video, which can detect new objects in a time-series image even if they stop moving after they enter the scene. Next, the CENTRIST feature [4] is applied to detect humans only in the foreground regions of each frame, to improve the computational efficiency.…”
Section: Basic Ideasmentioning
confidence: 99%
See 1 more Smart Citation
“…The first one is Normalized Vector Distance (NVD), for detecting differences in color, and the second one is Radial Reach Filter (RRF) [3], for detecting differences in texture. NVD is calculated as,…”
Section: Image Subtractionmentioning
confidence: 99%
“…Moreover, these techniques cannot cope with natural movements of the background, changing in the background geometry, or camera oscillations, resulting again in the generation of a great amount of false alarms. A partial solution to the issues raised by the previous discussion is given by the Radial Reach Filter algorithm (Satoh et al, 2002), where the thresholding operations to discern between background and foreground in the image are performed at pixel level, but taking into account a sort of local texture. One main concern remains: The fact that the threshold values are commonly chosen by the operator, implies that the performance of the algorithm is depending on the operator experience and skills.…”
Section: State Of the Artmentioning
confidence: 99%