An Embedded Monocular Vision Approach for Ground-Aware Objects Detection and Position Estimation
An Embedded Monocular Vision Approach for Ground-Aware Objects Detection and Position Estimation
复制标题
用于地面感知物体检测和位置估计的嵌入式单目视觉方法
DOI：
10.48550/arxiv.2207.09851
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Edna Barros
中科院分区：
文献类型：
--
作者：
João G. Melo;Edna Barros
在RoboCup小型联赛（SSL）中，鼓励团队提出解决方案，仅使用嵌入式传感信息在SSL场地内执行基本足球任务。因此，这项工作提出了一种嵌入式单目视觉方法，用于检测足球场内的物体和估计相对位置。通过假设地面上有物体，利用来自环境的先验知识，机载摄像机将其位置固定在机器人上。我们在NVIDIA Jetson Nano上实现了所提出的方法，并使用SSD MobileNet v2进行了2D目标检测，并进行了TensorRT优化，检测距离达3.5米的球、机器人和目标。球定位评估表明，该解决方案克服了目前使用的SSL视觉系统在距离车载摄像头1米以内的位置，均方根误差为14.37毫米。此外，该方法达到了平均30帧/秒的实时性。
In the RoboCup Small Size League (SSL), teams are encouraged to propose solutions for executing basic soccer tasks inside the SSL field using only embedded sensing information. Thus, this work proposes an embedded monocular vision approach for detecting objects and estimating relative positions inside the soccer field. Prior knowledge from the environment is exploited by assuming objects lay on the ground, and the onboard camera has its position fixed on the robot. We implemented the proposed method on an NVIDIA Jetson Nano and employed SSD MobileNet v2 for 2D Object Detection with TensorRT optimization, detecting balls, robots, and goals with distances up to 3.5 meters. Ball localization evaluation shows that the proposed solution overcomes the currently used SSL vision system for positions closer than 1 meter to the onboard camera with a Root Mean Square Error of 14.37 millimeters. In addition, the proposed method achieves real-time performance with an average processing speed of 30 frames per second.
DOI：
10.1109/34.888718
发表时间：
2000-11-01
影响因子：
23.6
作者：
Zhang, ZY
通讯作者：
Zhang, ZY