Study On Multiple-Camera Sub-Pixel Locating Algorithm Based On Correlation

Digital image measuring is a new measuring technology. It is based on the optics and composed of optics, electronics and computer. It syncretizes electronics, computer science, laser science, and image processing technology. It is used widely in the following fields such as geometry dimension measurement, aviation remote sensing measurement, high precision locating and face location about image.Edge location is always the focus point in the image measurement for the important status and effect in the project. In recent, sub-pixel edge location algorithm is dissatisfied in locating and testing for the target. For the reason, the high precision location measuring is researched according to the project.This thesis comes from a scientific research project named "LAMOST fiber location equipment" which belongs to a graduate school. The main task in the thesis is studying a kind of high precision location algorithm which will achieve high precision location in CCD and whose location precision hyper 0.01 pixel. So the multiple-cameras sub-pixel locating algorithm is brought forward based on correlation at simultaneity analyzing and discussing normal sub-pixel locating algorithm and it is analyzed how to detect the target template used in digital correlation. Another, improving on the pyramid method and deduce the scouting times.At last, the emulator experiment according to multiple-cameras sub-pixel locating algorithm is achieved based on correlation by MATLAB. The location precision hyper 0.01 pixel is testified by the algorithm and the demand of the project is satisfied according to the experiment result.