#### DMCA

## Recognition Of Planar Objects Over Complex Backgrounds Using Line Invariants And Relevance Measures (1995)

Venue: | Workshop on Geometric Modeling & Invariants for Computer Vision |

Citations: | 5 - 2 self |

### Citations

391 |
Object recognition by Computer: The role of Geometric Constraints.
- Grimson
- 1990
(Show Context)
Citation Context ...terize the shape [3]. The fact that primitives have already been extracted to compute invariant measures, makes invariantbased object recognition systems usually based on geometric hashing techniques =-=[6]-=-[11]. In this case, there are two major concerns. The first is that the indexing space must provide enough discrimination capability. The second is that the indexing function should remain simple enou... |

256 |
Geometric invariance in computer vision,
- Mundy, Zisserman
- 1992
(Show Context)
Citation Context ...rious points of view. 1. Introduction Geometric invariants are, in theory at least, valuable features to characterize a geometric structure because of their stability under various imaging conditions =-=[9]-=-. Given the knowledge of the applied camera transformation, the type of extracted features and the geometric structure of the analyzed shape, various invariants can be defined [8]. For example, a plan... |

154 |
Invariant descriptors for 3D object recognition and pose,”
- Forsyth
- 1991
(Show Context)
Citation Context ...ure of the analyzed shape, various invariants can be defined [8]. For example, a planar shape characterized by five points and examined through perspective projection may provide two invariant values =-=[5]-=-. Features such as lines [12] or corners [1] offer a more robust and precise positional information; line invariants therefore appear more suited for reliable indexing of objects from a model-base. Ot... |

47 |
Navigation using affine structure from motion
- Beardsley, Zisserman, et al.
- 1994
(Show Context)
Citation Context ...s can be defined [8]. For example, a planar shape characterized by five points and examined through perspective projection may provide two invariant values [5]. Features such as lines [12] or corners =-=[1]-=- offer a more robust and precise positional information; line invariants therefore appear more suited for reliable indexing of objects from a model-base. Other primitives like ellipses can be extracte... |

38 | Integration of bottom-up and top-down cues for visual attention using non-linear relaxation
- Milanese, Wechsler, et al.
- 1994
(Show Context)
Citation Context ...levance The work described in this paper is part of a larger project whose goal is to build an object recognition system, based on the concept of relevance [2] [10], on a focus-of-attention mechanism =-=[7]-=-, and on geometric hashing with invariant measures. The guiding idea is the transformation of a static image into several dataflows of image primitives (e.g. for line segments, for circular arcs, and ... |

20 |
An experimental evaluation of projective invariants
- Coelho, Heller, et al.
- 1992
(Show Context)
Citation Context ...bjects, with about 10 views for each object acquired on differing backgrounds, with differing orientations in 3D space. Each group of five coplanar lines can be characterized by two scalar invariants =-=[4]-=- [5], so for each polygon there are potentially possible five-lines combinations for which two invariant values can be evaluated. To avoid using such a high number of combinations for each shape repre... |

14 | Hierarchical object description using invariants
- Rothwell
- 1993
(Show Context)
Citation Context ...ize the shape [3]. The fact that primitives have already been extracted to compute invariant measures, makes invariantbased object recognition systems usually based on geometric hashing techniques [6]=-=[11]-=-. In this case, there are two major concerns. The first is that the indexing space must provide enough discrimination capability. The second is that the indexing function should remain simple enough s... |

14 |
Geometric Hashing With Line Features
- Tsai
- 1994
(Show Context)
Citation Context ...arious invariants can be defined [8]. For example, a planar shape characterized by five points and examined through perspective projection may provide two invariant values [5]. Features such as lines =-=[12]-=- or corners [1] offer a more robust and precise positional information; line invariants therefore appear more suited for reliable indexing of objects from a model-base. Other primitives like ellipses ... |

13 | Projectively Invariant Decomposition and Recognition of Planar Shapes
- Carlsson
- 1996
(Show Context)
Citation Context ...nal information; line invariants therefore appear more suited for reliable indexing of objects from a model-base. Other primitives like ellipses can be extracted (or fitted) to characterize the shape =-=[3]-=-. The fact that primitives have already been extracted to compute invariant measures, makes invariantbased object recognition systems usually based on geometric hashing techniques [6][11]. In this cas... |

8 |
Temporal precedence in asynchronous visual indexing
- Bost, Milanese, et al.
- 1993
(Show Context)
Citation Context ...base. 3. Extracting Line-invariants using Relevance The work described in this paper is part of a larger project whose goal is to build an object recognition system, based on the concept of relevance =-=[2]-=- [10], on a focus-of-attention mechanism [7], and on geometric hashing with invariant measures. The guiding idea is the transformation of a static image into several dataflows of image primitives (e.g... |

4 |
Active 3D object recognition using 3D affine invariants
- Vinther, Cipolla
- 1994
(Show Context)
Citation Context ...d is that the indexing function should remain simple enough so as not to excessively increase recognition time. This property would be of particular importance in applications such as object tracking =-=[13]. The pres-=-ent paper concentrates on following problems. Section 2. discusses the learning of invariant representations, based on a "five-lines" invariant, for automatic construction of object model. S... |

3 | Selecting relevant information and delaying irrelevant data for objects recognition
- Pun, Bost, et al.
- 1994
(Show Context)
Citation Context .... 3. Extracting Line-invariants using Relevance The work described in this paper is part of a larger project whose goal is to build an object recognition system, based on the concept of relevance [2] =-=[10]-=-, on a focus-of-attention mechanism [7], and on geometric hashing with invariant measures. The guiding idea is the transformation of a static image into several dataflows of image primitives (e.g. for... |

1 |
Projective geometry in computer vision", in: Handbook of Pattern Recognition and Computer
- Mohr
- 1993
(Show Context)
Citation Context ...s imaging conditions [9]. Given the knowledge of the applied camera transformation, the type of extracted features and the geometric structure of the analyzed shape, various invariants can be defined =-=[8]-=-. For example, a planar shape characterized by five points and examined through perspective projection may provide two invariant values [5]. Features such as lines [12] or corners [1] offer a more rob... |