創新及科技解決方案

解決方案編號

S-0001

解決方案名稱

PixelEye Object Detection

解決方案描述

PIXEL EYE, adopted TensorFlow and Keras technology, plus high-end GPU computing, which achieved real-time object detection, tracking and object category recognition. This technology can also be applied to crowd flow monitoring with providing different information about the flow of people: such as number of people, average dwell time, distribution of detected objects at each location, and predicted movement paths and traffic data.

It can achieve good accuracy even the scene is very crowded and also handle overlapping of people properly. Our tracking system can provide various information about the customers such as the number of people, their average staying time, the distribution at each location and predict the moving path and traffic data. Such information can be available in real time, and we provide customizable backend for managers to view and analyze such information at their convenience with processing high performance GPU.

應用領域

城市管理

工商業

發展

教育

就業及勞工

環境

財經

食物

衛生

房屋

基礎設施

法律及保安

人口

社會福利

運輸

使用的技術

人工智能

深度學習

機器學習

影像分析

使用例子

This technology can be used to replace traditional data statistics methods, and suitable for environments that are prolonged, dense, and heavily overlapping. Users can accurately grasp real-time detection/recognition results through Data Visualization reports. Which helps with security monitoring, multiple object/object tracking, market statistics, smart city planning, etc.

若政府部門欲對創科方案進行PoC試驗或技術測試,請聯絡Smart LAB。