Name
Abstract | ||
|---|---|---|
We introduce a new class of image features, the Ray fea- ture set, that consider image characteristics at distant con- tour points, capturing information which is difficult to rep- resent with standard feature sets. This property allows Ray features to efficiently and robustly recognize deformable or irregular shapes, such as cells in microscopic imagery. Ex- periments show Ray features clearly outperform other pow- erful features including Haar-like features and Histograms of Oriented Gradients when applied to detecting irregularly shaped neuron nuclei and mitochondria. Ray features can also provide important complementary information to Haar features for other tasks such as face detection, reducing the number of weak learners and computational cost. Ray features can be efficiently precomputed to reduce cost, just as precomputing integral images reduces the over- all cost of Haar features. While Rays are slightly more ex- pensive to precompute, their computational cost is less than that of Haar features for scanning an AdaBoost-based de- tector window across an image at run-time. |
Year | DOI | Venue |
|---|---|---|
2009 | 10.1109/ICCV.2009.5459210 | ICCV |
Keywords | Field | DocType |
Haar transforms,feature extraction,object detection,ray tracing,shape recognition,AdaBoost,Haar features,Ray feature set,face detection,image features | Computer vision,Object detection,Histogram,AdaBoost,Pattern recognition,Computer science,Ray tracing (graphics),Feature (computer vision),Feature extraction,Haar-like features,Artificial intelligence,Face detection | Conference |
Volume | Issue | ISSN |
2009 | 1 | 1550-5499 |
Citations | PageRank | References |
14 | 2.48 | 15 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Kevin Smith | 1 | 2430 | 88.78 |
| Alan Carleton | 2 | 18 | 3.63 |
| Vincent Lepetit | 3 | 6178 | 306.48 |