Description: 利用几何成像原理建立起CCD 双目立体视觉测量系统的数学模型,从提高系统测
量精度出发,在理论上重点对系统结构参数、图像识别误差与系统测量精度的关系进行了深入的分析和探讨,并通过实验对结论进行了验证。研究内容对实际建立该测量系统具有很强的指导作用。-Established the principle of the use of CCD imaging geometry binocular stereo vision measurement system of mathematical models to improve system measurement accuracy from the start, in theory, focus on structural parameters of the system, image recognition system error and the relationship between the measurement accuracy of an in-depth analysis and discussion and through experiments carried out to verify the conclusions. Study of the actual establishment of the measurement system has a strong guiding role. Platform: |
Size: 339968 |
Author:chenruibao |
Hits:
Description: For fast image processing this script extracts all the parameters for stereo image rectification in order to get the largest rectangular image pairs with only valid pixel values. The parameters are sorted in the way images are read in (First pixels of first row, then pixels of 2nd row etc.). Furthermore the precision for the bilinear transformation can be reduced to 8, 16, 32 or nearest neighbor in order to save calculation efforts. The effect of precision reduction are evaluated with Double_versus_uint.m .
Requirements:
This script works only in conjunction with the Camera Calibration Toolbox for Matlab from Jean-Yves Bouguet:
http://www.vision.caltech.edu/bouguetj/calib_doc/
Also required is the modified file FindLargestRectanglesPUS.m (original from Jaroslaw Tuszynski, see
http://www.mathworks.com/matlabcentral/fileexchange/28155-inscribed-rectangle/content/FindLargestRectangles.m)
This files is included in this package. - For fast image processing this script extracts all the parameters for stereo image rectification in order to get the largest rectangular image pairs with only valid pixel values. The parameters are sorted in the way images are read in (First pixels of first row, then pixels of 2nd row etc.). Furthermore the precision for the bilinear transformation can be reduced to 8, 16, 32 or nearest neighbor in order to save calculation efforts. The effect of precision reduction are evaluated with Double_versus_uint.m .
Requirements:
This script works only in conjunction with the Camera Calibration Toolbox for Matlab from Jean-Yves Bouguet:
http://www.vision.caltech.edu/bouguetj/calib_doc/
Also required is the modified file FindLargestRectanglesPUS.m (original from Jaroslaw Tuszynski, see
http://www.mathworks.com/matlabcentral/fileexchange/28155-inscribed-rectangle/content/FindLargestRectangles.m)
This files is included in this package. Platform: |
Size: 116736 |
Author:Adilson |
Hits:
Search - image rectification in matlab