基于双目立体视觉的海浪波面三维重建技术

为了提高海面特征点检测的准确度和三维重建的精度,在基于传统的Harris算法的基础上,提出1种基于高斯金字塔图像的改进Harris特征点检测算法.利用搭建的双目相机平台,对海浪图像进行采集并完成相机的标定过程,然后根据改进的角点提取算法对图像的角点进行检测,利用尺度不变特征变换(scale-invariant feature transform,SIFT)算法对海浪图像特征点进行立体匹配得出视差图,最后根据三角测量原理获取图像的深度信息,实现海浪波面的三维信息重建.实验结果证明,在针对海浪图像时,该方法具有更高的精度和准确度.

Three-dimensional Reconstruction of Wave Surface Based on Binocular Stereo Vision

In order to improve the accuracy of sea surface feature point detection and the accuracy of 3D reconstruction, on the basis of traditional Harris algorithm, an improved Harris feature point detection algorithm based on Gaussian pyramid image was proposed. Using the built binocular camera platform, the sea wave image was collected and the camera calibration process was completed, then the corner points of the image were detected according to the improved corner point extraction algorithm, and the stereo image matching of the sea wave image feature points were obtained by using the SIFT algorithm to obtain parallax Figure, and finally obtain the depth information of the image according to the principle of triangulation to realize the reconstruction of the three-dimensional information of the wave surface. The experimental results prove that this method has higher precision and accuracy when targeting the sea wave image.