Research Of Camera Calibration And3D Reconstruction

3D reconstruction has become an important area in the field of computer vision research. Recently, image-based3D reconstruction techniques gradually become a research focus in this field. The advantage of this method is it has lower time complexity and can realize semi-automatic or even automatic modeling. Because of its advantages, this technique has been widely used in mobile robot navigation systems, aviation, medical imaging, industrial automation systems and other various fields.In this paper, the multiple view images of an object are used to achieve the3D reconstruction. We do the research of reconstruction mainly in the image processing, feature extraction and matching, camera calibration,3D reconstruction and optimization. This paper will be divided into the following sections:Image processing. A new image segmentation method based on fractal dimension has been presented in this section, compared with other methods the advantage of this method is it improves the accuracy of the segmentation result and also can improve the efficiency. A fractal edge detection method is also introduced in this section, the experiment result has proved that the fractal edge feature has better anti-noise performance.Feature detection and correspondence. Several feature point detection methods and the SIFT algorithm has been studied in this section, and feature correspondence between two images has been realized in experiment section.Camera calibration. The principles of Chart-Based Calibration are introduced in this section, and then a concentric circle chart has been used to achieve the camera calibration. This method automatically detect the points of the pattern features, the experimental results show the robustness and the accuracy.3D reconstruction. The optimization process which has be used in our paper has been introduced in this section. The main method is using the initial estimated camera parameter, the result of the MVS on downscaled images, and the visual information and photometric consistency of the patch, and finally combining with the SBA software, after several iterations to get the final results of the optimization of3D reconstruction.