基于双目融合的场馆深度信息视觉交互仿真

Visual Interaction Simulation of Depth Information of Stadium Based on Binocular Fusion

采用目前方法对场馆深度信息进行交互时,由于未结合双目立体视觉结构获取目标物体的原始图像数据,使得最终获取的信息交互结果存在信息交互可信度较差、信息交互往返时延较长的问题。提出基于双目融合的场馆深度信息视觉交互仿真的方法,结合双目立体视觉结构获取目标物体同一水平不同位点的两组原始图像数据,并对原始图像数据进行预处理,获取基于优化数据的特征角点匹配图像,将征角点已匹配的目标物体图像输入以大数据背景下神经网络为基础建立的信息交互模型中,得到目标物体的三维几何深度信息,同时实现信息的交互。仿真结果表明,所提方法信息交互可信度较高、信息交互往返时延较短。

When using current methods to interact with depth information in venues,the lack of integration with binocular stereo vision structures to obtain the original image data of the target object results in poor information inter-action credibility and long round-trip information interaction latency in the final information interaction results.A method for visual interactive simulation of depth information in venues based on binocular fusion is proposed.Two sets of raw image data of the same horizontal position of the target object are acquired using binocular stereo vision struc-ture,and the raw image data is preprocessed to obtain feature corner matching images based on optimized data.The target object image that has been matched by the feature corners is input into an information interaction model based on neural networks in the context of big data,and the 3D geometric depth information of the target object is obtained,while information interaction is realized.The simulation results show that the proposed method has higher reliability of information interaction and shorter round-trip delay of information interaction.