Digital Signal Processing Reference
In-Depth Information
Calibration System for the Cross-Correlation Algorithm
Based on OpenCV
*
Min Chen and Hongrong Wang
Fujian University of Technology,
Department of Computer and Information Science
Fuzhou, PRC China
chenmin@fjut.edu.cn
Abstract
.
The application of cross-correlation algorithm should be preceded by
the calibration system validating the precision and accuracy of the algorithm.
Selected adjacent two images from the video stream of the moving target, and
calculate the displacement of the moving target, and then the vector information
can be obtained according to the time interval set in the video capture rate.
Cross-correlation algorithm in order to reduce computation, the image on the
equivalent land is divided into several regions, the sub-regions of the same loca-
tion is calculated. The selection of the size of the sub-regions will affect the ac-
curacy of measurement results, so the image calibration system using a variety
of images to find the association between the size of the sub-regions and the in-
tensity of the images. This will reduce the generation of error, and founded a
basement for the subsequent processing. Because the selection of the sub-
regions size will affect the accuracy of measurement results, the image calibra-
tion system testing images with variety intensities to find the association
between the size of the sub-regions and the image intensities. This will reduce
the generation of error, and lay a good foundation for the subsequent
processing. The development of the image calibration system is developed un-
der Windows, using C/C++ programming languages, and the openCV open
source image library.
Keywords:
calibration system, cross-correlation algorithm, sub-region,
openCV
1
Introduction
Cross-correlation technique has been successfully applied in many fields[1]; its de-
velopment was based on information theory and stochastic process theory. In the
cross-correlation algorithm, the image of the moving target is a random distribution of
the digital field, using the correlation search technology to extract the displacement
*
This research was supported by the general project of the Natural Science Foundation of Fu-
jian Province (No. 2008F3003), General project of the Fujian Province Department of Educa-
tion Technology (No. JB08193), research project of Fujian University of Technology (No.
GY-Z0803).
