基于姿态传感器进行感兴趣区域压缩的园区无人车图传

Image transmission of an automated guided vehicle in industrial environments that compressed by region of interest based on the attitude sensor

摘要:

园区无人车需要同时用于视觉导航、监控及控制信号传输的带宽.在实时提供图像信号的同时,不过多占用带宽,成为园区无人车图传的关键.而用于导航监控的图像要求水平范围视场具有较高的图像质量，但车辆行驶过程中路面不平整，使得水平视场的获取呈现一定难度.过多的带宽占用会影响无人车视觉导航、监控及控制信号的传输，因此,在低带宽条件下实现图像数据传输的实时性是无人车图传研究的重点.为了平衡图像质量与带宽占用，本文提出基于H.264/AVC算法，利用姿态传感器实现以水平视场为感兴趣区域的图像压缩算法，对视频帧分为感兴趣区域和非感兴趣区域采用不同的压缩比压缩.相比H.264/AVC算法，此方法在保证图像质量的情况下，能够实现更高的压缩率，对网络带宽更友好.

Abstract:

The bandwidth of an automated guided vehicle in industrial environments should meet the requirements of visual navigation, monitoring and control signal transmission simultaneously. Providing image signals in real time while not consuming too much bandwidth has become the key to the transmission of automated guided vehicle in industrial environments. The images used for navigation monitoring should be of high quality in the horizontal field of view. However, the acquisition of the view in the horizontal field is difficult because the road surface is uneven during the moving of vehicles, and the excessive bandwidth occupation of view acquisition will affect the navigation, monitoring and control signal transmission of the automated guided vehicle. Therefore, transmitting image data in real time under lowbandwidth conditions is the focus of automated guided vehicle image transmission research. The image used for navigation monitoring should be of high quality in the horizontal field of view. In order to balance image quality against bandwidth consumption, this paper proposes an image compression algorithm based on the H.264/AVC algorithm. The proposed algorithm uses a gesture sensor to obtain the horizontal field of view as the region of interest, and adopts different compression rates for the region of interest and the region of no interest in the video frame, respectively. Compared to the H.264/AVC algorithm, this method can achieve a higher compression rate while ensuring the image quality, and is more friendly to the network bandwidth.