无人机
计算机科学
计算机视觉
人工智能
事件(粒子物理)
船上
图像(数学)
融合
实时计算
语言学
哲学
物理
遗传学
量子力学
工程类
生物
航空航天工程
作者
Xinjun Cai,Jingao Xu,Kuntian Deng,Hongbo Lan,Yue Wu,Xiangwen Zhuge,Zheng Yang
出处
期刊:ACM Transactions on Sensor Networks
[Association for Computing Machinery]
日期:2024-09-20
摘要
Drones have witnessed extensive popularity among diverse smart applications, and visual SLAM technology is commonly used to estimate 6-DoF pose for the drone flight control system. However, traditional image-based SLAM cannot ensure the flight safety of drones, especially in challenging environments such as high-speed flight and high dynamic range scenarios. Event camera, a new vision sensor, holds the potential to enable drones to overcome the above challenging scenarios if fused into the image-based SLAM. Unfortunately, the computational demands of event-image fusion SLAM have grown manifold compared to image-based SLAM. Existing research on visual SLAM acceleration cannot achieve real-time operation of event-image fusion SLAM on on-board computing platforms for drones. To fill this gap, we present TrinitySLAM , a high accuracy, real-time, low energy consumption event-image fusion SLAM acceleration framework utilizing Xilinx Zynq, an on-board heterogeneous computing platform. The key innovations of TrinitySLAM include a fine-grained computation allocation strategy, several novel hardware-software co-acceleration designs, and an efficient data exchange mechanism. We fully implement TrinitySLAM on the latest Zynq UltraScale+ platform and evaluate its performance under one self-made drone dataset and four official datasets covering various scenarios. Comprehensive experiments show TrinitySLAM improves the pose estimation accuracy by 28% with half end-to-end latency and 1.2 × energy consumption reduction, compared to the most comparable SOTA heterogeneous computing platform acceleration baseline.
科研通智能强力驱动
Strongly Powered by AbleSci AI