视觉传感网络图像数据曲面重构方法仿真  

Simulation of Image Data Surface Reconstruction Method Based on Vision Sensor Network

作　　者：徐亦丹[1] 葛菁[1] XU Yi-dan;GE Jing(Institute of Technology,East China Jiao Tong University,Nanchang Jiangxi 330100,China)

机构地区：[1]华东交通大学理工学院,江西南昌330100

出　　处：《计算机仿真》2020年第5期473-476,481,共5页Computer Simulation

基　　金：江西省高等学校教学改革课题编号(JXJG-17-35-5)。

摘　　要：由于传统方法构成曲面模型的特征量较大,因此在算法处理与坐标表达上较为复杂,重构后的图像细节数据缺失,且重构时间较长。为解决上述问题,提出视觉传感网络图像数据曲面重构。首先利用CCD摄像机获取原始图像的坐标信息,并根据公式推算出图像曲面重构后任意一点的三维坐标;其次根据自适应采样策略对坐标数据做预处理,进而描绘出一条最小二乘意义下曲线,最后采用NURBS与Bezier相结合的方法计算图像曲面差值,通过改变控制顶点,实现图像数据的重构。仿真结果表明,曲面重构模型的表面非常光滑,特征参数近似服从正态分布,重构精准度以及重构效率有所提高,适合视觉传感网络图像数据的重构。In traditional methods, the algorithm and coordinate expression are too complex, and the detail data of reconstructed image are missing. Therefore, a surface reconstruction method for image data in vision sensor network was proposed. Firstly, the coordinate information of original image was obtained by CCD camera, and the 3 D coordinate of any point after the reconstruction of image surface was calculated by the formula. Secondly, the coordinate data was preprocessed according to the adaptive sampling strategy, and then a curve in the least square sense was drawn. Finally, the difference value of image surfaces was calculated by combining with NURBS and Bezier. After the control vertex was changed, we achieved the reconstruction of image data. Simulation results show that the surface of curved surface reconstruction model is very smooth. The feature parameters are approximately normal distribution. The reconstruction accuracy and reconstruction efficiency are improved. The proposed method is suitable for the reconstruction of image data of vision sensor network.

关 键 词：视觉传感网络 图像数据 曲面重构 数据处理 

分 类 号：TP318.2[自动化与计算机技术—计算机软件与理论]