#### DMCA

## Hierarchical Disparity Estimation with Programmable 3D Hardware (2004)

Venue: | Hardware”, WSCG (International Conference in Central Europe on Computer Graphics, Visualization and Computer Vision), Short Communications |

Citations: | 11 - 0 self |

### Citations

324 | Linear algebra operators for gpu implementation of numerical algorithms”,
- Kurger, Westermann
- 2003
(Show Context)
Citation Context ... programmability and computational power of recent 3D hardware is used for non-graphical purposes as well, e.g. for numerical calculations and simulations ([Hopf99b], [Hopf99a], [Harri02], [Thomp02], =-=[Krü03]-=-, [Bolz03]). Programmable graphics hardware is used in the field of image processing for linear and nonlinear filtering ([Hadwi01], [Yang02a], [Sugit03], [Colan03], [Viola03]). Some authors present pe... |

297 | Sparse Matrix Solvers on the GPU: Conjugate Gradients and Multigrid”.
- Bolz, Farmer, et al.
- 2003
(Show Context)
Citation Context ...ability and computational power of recent 3D hardware is used for non-graphical purposes as well, e.g. for numerical calculations and simulations ([Hopf99b], [Hopf99a], [Harri02], [Thomp02], [Krü03],=-= [Bolz03-=-]). Programmable graphics hardware is used in the field of image processing for linear and nonlinear filtering ([Hadwi01], [Yang02a], [Sugit03], [Colan03], [Viola03]). Some authors present performance... |

193 | A userprogrammable vertex engine.
- Lindholm, Kilgard, et al.
- 2001
(Show Context)
Citation Context ...dentifies and discards false matches. 2 RELATED WORK The demand for more complex and more realistic virtual realities has accelerated the development of highly programmable 3D graphics hardware (e.g. =-=[Lindh01]-=-). The programmability and computational power of recent 3D hardware is used for non-graphical purposes as well, e.g. for numerical calculations and simulations ([Hopf99b], [Hopf99a], [Harri02], [Thom... |

132 | Multi-resolution real-time stereo on commodity graphics hardware,” in
- Yang, Pollefeys
- 2003
(Show Context)
Citation Context ...sired depth resolution. Later, Yang and Pollefeys refined the method to work even for two camera setups using mipmapping facilities of graphics hardware to calculate a more robust similarity function =-=[Yang03]-=-. Sugita et al. [Sugit03] compare the performance of CPU and GPU implementations of several image filters and implemented a rather simple stereo matching algorithm. Their method is a blockbased matchi... |

103 | Physically-based visual simulation on graphics hardware.
- Harris, Coombe, et al.
- 2002
(Show Context)
Citation Context ... (e.g. [Lindh01]). The programmability and computational power of recent 3D hardware is used for non-graphical purposes as well, e.g. for numerical calculations and simulations ([Hopf99b], [Hopf99a], =-=[Harri02],-=- [Thomp02], [Krü03], [Bolz03]). Programmable graphics hardware is used in the field of image processing for linear and nonlinear filtering ([Hadwi01], [Yang02a], [Sugit03], [Colan03], [Viola03]). Som... |

93 | Detecting binocular half-occlusions: Empirical comparisons of five approaches.
- Egnal, Wildes
- 2002
(Show Context)
Citation Context ...n 3.5.2). 3.4 Disparity Verification The role of the input images can be exchanged and the matching procedure can be applied to the reversed pair of images (back-matching, bidirectional matching, see =-=[Egnal02]-=- for a comparison of disparity verification methods). In addition to thesleft-to-right disparity map DL a map DR of rightto-left disparities is obtained. We discard the disparity value for pixel p, if... |

76 | Using modern graphics architectures for general-purpose computing: a framework and analysis. In: - Thompson, Hahn, et al. - 2002 |

66 | Calculating dense disparity maps from color stereo images, an efficient implementation”,,
- Muhlmann, Maier, et al.
- 2002
(Show Context)
Citation Context ...inal pixel (within some allowed threshold). Because of the hierarchical approach we evaluate the disparity range at sparse positions and the optimized right-to-left disparity calculation described in =-=[Müh02]-=- cannot be applied. In our framework right-to-left disparity estimation is completely independent from left-to-right disparity calculation and requires the same amount of operations. 3.5 Optimizations... |

49 | Accelerating 3D Convolution Using Graphics Hardware Case Study,”
- Hopf, Ertl
- 1999
(Show Context)
Citation Context ...cs hardware (e.g. [Lindh01]). The programmability and computational power of recent 3D hardware is used for non-graphical purposes as well, e.g. for numerical calculations and simulations ([Hopf99b], =-=[Hopf99a],-=- [Harri02], [Thomp02], [Krü03], [Bolz03]). Programmable graphics hardware is used in the field of image processing for linear and nonlinear filtering ([Hadwi01], [Yang02a], [Sugit03], [Colan03], [Vio... |

41 | Fast Image Segmentation and Smoothing Using Commodity Graphics Hardware.”
- Yang, Welch
- 2002
(Show Context)
Citation Context ...and simulations ([Hopf99b], [Hopf99a], [Harri02], [Thomp02], [Krü03], [Bolz03]). Programmable graphics hardware is used in the field of image processing for linear and nonlinear filtering ([Hadwi01],=-= [Yang02a],-=- [Sugit03], [Colan03], [Viola03]). Some authors present performance comparisons between optimized CPU and GPU implementations of image processing methods. In general, if the GPU’s facilities for par... |

23 | Fast and accurate color image processing using 3-D graphics cards
- Colantoni, Boukala, et al.
(Show Context)
Citation Context ...99b], [Hopf99a], [Harri02], [Thomp02], [Krü03], [Bolz03]). Programmable graphics hardware is used in the field of image processing for linear and nonlinear filtering ([Hadwi01], [Yang02a], [Sugit03],=-= [Colan03],-=- [Viola03]). Some authors present performance comparisons between optimized CPU and GPU implementations of image processing methods. In general, if the GPU’s facilities for parallel computation and ... |

23 | Hardware-based nonlinear filtering and segmentation using high-level shading languages
- Viola, Kanitsar, et al.
- 2003
(Show Context)
Citation Context ...99a], [Harri02], [Thomp02], [Krü03], [Bolz03]). Programmable graphics hardware is used in the field of image processing for linear and nonlinear filtering ([Hadwi01], [Yang02a], [Sugit03], [Colan03],=-= [Viola03])-=-. Some authors present performance comparisons between optimized CPU and GPU implementations of image processing methods. In general, if the GPU’s facilities for parallel computation and fast textur... |

15 | Hardware-accelerated high-quality filtering on PC hardware
- Hadwiger, Theußl, et al.
- 2001
(Show Context)
Citation Context ...lculations and simulations ([Hopf99b], [Hopf99a], [Harri02], [Thomp02], [Krü03], [Bolz03]). Programmable graphics hardware is used in the field of image processing for linear and nonlinear filtering =-=([Hadwi01],-=- [Yang02a], [Sugit03], [Colan03], [Viola03]). Some authors present performance comparisons between optimized CPU and GPU implementations of image processing methods. In general, if the GPU’s facilit... |

14 | Accurate Dense Stereo Reconstruction using Graphics Hardware
- C, Klaus, et al.
- 2003
(Show Context)
Citation Context ... as input. The GPU implementation of the stereo matcher runs about twice as fast as the optimized CPU implementation using the Intel Performance Primitives library. This work extends our earlier work =-=[Zach03a]-=- in several aspects. Our previous implementation iteratively refines the vertices of the current 3D mesh hypothesis and keeps track of the best local modification so far. Evaluating the current mesh h... |

7 | Optimized stereo reconstruction using 3d graphics hardware
- Zach, Klaus, et al.
- 2003
(Show Context)
Citation Context ...odification so far. Evaluating the current mesh hypothesis requires rendering a large mesh, which decreases the overall performance. In later work we improved the speed of the algorithm significantly =-=[Zach03b]-=-, but the basic idea remained the same. In contrast to our current method the previous one calculates depth (or disparity) values only for mesh vertices, which are placed every 4 pixels. We still util... |

6 |
Hardware-Accelerated Wavelet Transformations
- Hopf, Ertl
- 2000
(Show Context)
Citation Context ...e 3D graphics hardware (e.g. [Lindh01]). The programmability and computational power of recent 3D hardware is used for non-graphical purposes as well, e.g. for numerical calculations and simulations (=-=[Hopf99b],-=- [Hopf99a], [Harri02], [Thomp02], [Krü03], [Bolz03]). Programmable graphics hardware is used in the field of image processing for linear and nonlinear filtering ([Hadwi01], [Yang02a], [Sugit03], [Col... |

6 |
Real-time consensus based scene reconstruction using commodity graphics hardware
- Yang, Welch, et al.
- 2002
(Show Context)
Citation Context ...ds. In general, if the GPU’s facilities for parallel computation and fast texture access can be exploited, GPU based algorithms can be several times faster than the CPU based counterpart. Yang et al=-=. [Yang02b]-=- present a real-time depth estimation method utilizing graphics hardware working with several calibrated cameras to increase robustness. Depth values are obtained by a plane sweep through the 3D world... |

5 |
Performance evaluation of programmable graphics hardware for image filtering and stereo matching
- Sugita, Naemura, et al.
- 2003
(Show Context)
Citation Context ...ions ([Hopf99b], [Hopf99a], [Harri02], [Thomp02], [Krü03], [Bolz03]). Programmable graphics hardware is used in the field of image processing for linear and nonlinear filtering ([Hadwi01], [Yang02a],=-= [Sugit03],-=- [Colan03], [Viola03]). Some authors present performance comparisons between optimized CPU and GPU implementations of image processing methods. In general, if the GPU’s facilities for parallel compu... |