一种广角–变焦协同的视觉感知装置应用设计
Application Design of a Wide-Angle and Zoom Collaborative Visual Perception Device
DOI: 10.12677/iae.2026.141004, PDF,    国家自然科学基金支持
作者: 刘德洋:南京工程学院创新创业学院,江苏 南京;罗中华, 阮英兰:南昌理工学院电子信息学院,江西 南昌;沈克永:南昌理工学院计算机信息工程学院,江西 南昌;徐梦溪, 郑胜男:南京工程学院计算机工程学院,江苏 南京
关键词: 城市物流视觉感知双镜头成像智能计算轻量化技术Urban Logistics Visual Perception Dual-Lens Imaging Intelligent Computing Lightweight Technology
摘要: 针对视频监控中全景覆盖与细节捕捉难以兼顾、计算资源及成本受限、带宽限制与数据传输滞后,以及对广域分散部署的固定或车载移动视频监控复杂场景适应性不足等问题,本文提出一种广角–变焦协同的视觉感知装置(简称:WAZC-VPD)应用设计方案。该方案通过创新设计广角 + 变焦双镜头协同成像组件与三核异构轻量化CPU架构,集成轻量化Smart侦测技术及5G通信模块,构建“双镜成像 + 三核计算 + 5G传输”一体化轻量化架构,实现了“全景覆盖 + 细节捕捉”双模式联动的视觉智能感知及视频数据5G实时传输。结合城市物流视觉感知应用场景的现场测试结果也表明,WAZC-VPD具备合理性与有效性,为城市物流视频监控场景提供了一种可选的先进视觉感知手段。
Abstract: Aiming at the problems in video surveillance, such as the difficulty in balancing panoramic coverage and detail capture, limited computing resources and costs, bandwidth constraints and data transmission delay, as well as insufficient adaptability to complex scenarios of wide-area distributed fixed or vehicle-mounted mobile video surveillance, this paper proposes an application design scheme for a Wide-Angle and Zoom Collaborative Visual Perception Device (abbreviation: WAZC-VPD). By innovatively designing a wide-angle + zoom dual-lens collaborative imaging module and a triple-core heterogeneous lightweight CPU architecture, integrating lightweight Smart detection technology and 5G communication module, the scheme constructs an integrated lightweight architecture of “dual-lens imaging + triple-core computing + 5G transmission”. This architecture enables visual intelligent perception with dual-mode linkage of “panoramic coverage + detail capture” and real-time 5G transmission of video data. Field test results combined with the application scenario of urban logistics visual perception show that the WAZC-VPD is reasonable and effective, providing an optional advanced visual perception method for the urban logistics video surveillance scenario.
文章引用:刘德洋, 罗中华, 沈克永, 阮英兰, 徐梦溪, 郑胜男. 一种广角–变焦协同的视觉感知装置应用设计[J]. 仪器与设备, 2026, 14(1): 19-30. https://doi.org/10.12677/iae.2026.141004

参考文献

[1] Zhou, W., Zhao, L., Zhang, R.Y., Cui, Y.F., Huang, H.P., Qie, K. and Wang, C. (2024) Vision Technologies with Applications in Traffic Surveillance Systems: A Holistic Survey. arXiv: 2412.00348.
[2] 华经产业研究院. 2025-2031年中国物联网安防行业发展前景预测及投资方向研究报告[EB/OL].
https://www.huaon.com/channel/security/1076702.html, 2025-07-18.
[3] 湖南贝哲斯信息咨询有限公司. 2023年全球与中国人工智能视频监控行业前景预测报告[EB/OL].
https://www.shangyexinzhi.com/article/9990378.html, 2023-07-18.
[4] Lu, Y., Wang, Z., Liu, M., Wang, H. and Wang, L. (2023) Learning Spatial-Temporal Implicit Neural Representations for Event-Guided Video Super-Resolution. 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, 17-24 June 2023, 1557-1567. [Google Scholar] [CrossRef
[5] 徐梦溪, 沈克永, 涂宏斌, 等. 一种视频与物流业务数据融合的效率-安全双目标协同优化系统设计模式[J]. 软件工程与应用, 2025, 14(4): 897-905.
[6] 徐梦溪, 刘姝怡, 程晓玲, 等. 基于多光谱成像与边缘计算的物流安全风险预警模式及系统实现[J]. 计算机科学与应用, 2025, 15(10): 85-96.
[7] 曹行健, 张志涛, 孙彦赞, 等. 面向智慧交通的图像处理与边缘计算[J]. 中国图象图形学报, 2022, 27(6): 1743-1767.
[8] Cheng, X., Xu, M., Yan, X., Yang, Y., Xu, Y. and Ruan, Y. (2024) A Design Pattern of IAPVS Platform Based on Distributed Edge Computing. Journal of Physics: Conference Series, 2732, Article ID: 012001. [Google Scholar] [CrossRef
[9] Zhang, X., Yoon, J., Bansal, M. and Yao, H. (2024) Multimodal Representation Learning by Alternating Unimodal Adaptation. 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, 16-22 June 2024, 27446-27456. [Google Scholar] [CrossRef
[10] Zou, Z., Chen, K., Shi, Z., Guo, Y. and Ye, J. (2023) Object Detection in 20 Years: A Survey. Proceedings of the IEEE, 111, 257-276. [Google Scholar] [CrossRef
[11] 王晨, 周威, 章世祥. 一种特征融合的视频事故快速检测方法[J]. 交通运输工程与信息学报, 2022, 20(1): 31-38.
[12] 徐梦溪, 罗中华, 程晓玲, 等. 基于双镜头视野协同成像的无线视频传感器网络构建[J]. 传感器技术与应用, 2024, 12(1): 54-62.
[13] 徐梦溪, 刘姝怡, 刘梓莹, 等. 一种基于DVFS特性曲线的异构计算单元低功耗协同调度方法[J]. 软件工程与应用, 2025, 14(6): 1219-1230.