Using the DragonBoard™ 410c from Arrow Electronics, this project is designed to acquire images of the surrounding environment using the camera, store the images, and use special image processing and recognition algorithms to scan and determine if there is fire in the captured images to achieve flame recognition.
The project was inspired by previous exposure to flame recognition, where a STM32F749 chip of low frequency was used and the realization effect was not very good. This project is an effort to continue to realize this function on a high speed embedded development platform. The desired objective is to make the recognition of small area fire achieve accuracy of 90%.
- Creator: Li Ning, Student of Taiyuan University of Science and Technology
- Project Name: Flame Recognition
- Type of Project: Demo
- Project Category: IoT, Smart Home
- Board(s) used: DragonBoard 410c
- View project on Qualcomm Developer Network
- New flame detection algorithm based on color characteristicshttp://xuebao.jlu.edu.cn/gxb/CN/Y2014/V44/I6/1787
- High precision video flame detection algorithm based on multi-characteristics infusionhttp://xuebao.jlu.edu.cn/gxb/CN/abstract/abstract10105.shtml