結合光達與穿戴裝置之智慧監控

Integrating LiDAR and Wearable Devices for Smart Surveillance

邱彥豪(Yen-Hao Chiu)

指導教授 ： 曾煜棋

國立交通大學/資訊學院/網路工程研究所/碩士(2019年)

摘要

在日常生活中，重要區域有安全監控的需求。透過電腦視覺的技術，可得知影像畫面中的人員資訊。若要進一步辨識人員身分，則需人員的生物特徵，如：虹膜、指紋、人臉等。但攝影機的畫面容易受到光線、拍攝角度的影響，需面對攝影機才可正確地識別。透過使用者身上的穿戴式裝置，如RFID tag、穿戴式手環、智慧型手機等，亦可作為人員的識別憑證。近年來，隱私權的議題越來越受到重視，在不使用影像資訊的前提下，如何得知環境中的人員資訊變成一項重要挑戰。在自駕車與機器人的情境裡，光學雷達(LiDAR，亦稱光達)常被用於掃描周圍環境、辨識環境中的障礙物以避免碰撞。目前市面上的產品已有2D與3D LiDAR兩種類型，3D LiDAR 雖然可以提供高解析度的點雲影像但3D LiDAR的成本過高，很難普及應用在日常生活中。在本論文中，我們整合較低成本的2D LiDAR以及穿戴式裝置，掃描環境中是否有人員出現，並且識別他們的身分。即使在低光源的環境下，我們提出的系統可分析人員之活動模式，在LiDAR的掃描結果顯示其身分資訊。即使沒有生物特徵仍可辨識出人員身分。此系統包含短期配對與長期配對的整合機制，可更有效地識別人員身分。經由實驗評估，我們的系統可提供92%的人員識別精確度。

關鍵字

資料融合 ； 物聯網 ； 光達 ； 身分識別 ； 穿戴式計算

並列摘要

Surveillance has been widely used for security and management purposes. Through computer vision technologies, we can capture human and other objects. In order to further recognize the identity of a person, biological characteristics, such as iris, fingerprint, face, etc., are needed. But image recognition usually sets a strong restriction in recognition scenario, image resolution, light condition and shooting angle. In this work, we are interested in using Light Detection and Ranging (LiDAR), which is often used in environment and obstacle scanning, for person identification. Unlike camera, LiDAR cannot capture biological features. We further integrate wearable devices, such as RFID badge or smart watch, which are virtually the personal identity nowadays in many application scenarios, to achieve personal identification. We present our multi-sensory data fusion algorithm, as well as system prototyping, which integrates a low-cost 2D LiDAR and some wearable devices. The advantages of our approach are: (i) it works in low-light conditions, (ii) it relieves the privacy concern that is often brought up in camera-based solutions, and (iii) it helps visualize personal profiles easily in LiDAR scans. We demonstrate several smart surveillance applications, such as location tracking, smart fencing, and intrusion detection.

並列關鍵字

Data Fusion ； IoT ； LiDAR ； Surveillance ； Wearable Computing

參考文獻

[1] R. T. Collins et al., “A system for video surveillance and monitoring,” VSAM final report, pp. 1–68, 2000.

[2] X. Wang, “Intelligent multi-camera video surveillance: A review,” Pattern Recognition Letters, vol. 34, no. 1, pp. 3–19, 2013.

[3] W. Zhao, R. Chellappa, P. J. Phillips, and A. Rosenfeld, “Face recognition: A literature survey,” ACM Computing Surveys, vol. 35, no. 4, pp. 399–458, 2003.

[4] L. Hong, Y. Wan, and A. Jain, “Fingerprint image enhancement: algorithm and performance evaluation,” IEEE Transactions on Pattern Analysis and Machine Intelligence,

vol. 20, no. 8, pp. 777–789, 1998.