山区智能视觉火情高空图像搜集系统的设计与实现

介绍了一种针对山区多样性突发火情事件中基于智能视觉的高空图像搜集系统,采用以ST意法半导体公司的STM32F107VCT6为处理核心的开发平台,地面控制平台通过高空无人机的高速照相机对目标区域火情图像数据进行搜集,并将图像数据带回到地面控制平台的PC处理器上进行综合分析。地面控制平台根据智能视觉信息采集模块采集的山区遥感图像的均值作为初始背景,并对背景图像进行实时更新,利用光谱特征与距离算法获取差分图像,并进行图像处理,从而获取山区发生火情区域的图像分析数据,以实现对山区森林突发火情事件的及时监测,在火情发生初始阶段就能扑灭。本文对山区智能视觉火情高空搜集系统的实现原理、核心硬件设计、系统软件设计等分别进行了研究,并通过系统实验进行了验证,测试结果表明,该系统在对山区的高空图像搜集实验中,对不同颜色物体的分别识别率为100%,对火灾与地形地貌的识别率为95.55%和81.12%,非常适用于各种山区环境下的智能视觉监控系统。

Design and implementation of intelligent visual fire mountain altitude image collection system

ntroduce a sudden fire incident in the high-altitude mountain biodiversity intelligent vision-based image acquisition system, using the ST STMicroelectronics's STM32F107VCT6 to deal with the core development platform, high-altitude UAV ground control platform via a high-speed camera fire on the target area to collect image data and image data back to ground control platform of PC processors were analyzed. Ground control platform based on the average intelligent visual information acquisition module mountainous backdrop as the initial remote sensing images, background images and real-time updates, the use of spectral characteristics and distance algorithm to obtain the difference image, and image processing to get the fire occurred mountainous region the image analysis data in order to achieve the mountainous forest fire emergencies timely monitoring of events occurred in the initial stages of the fire can be extinguished. In this paper, fire-altitude mountain intelligent vision system to achieve the principle of collecting core hardware design, software design and other studies were carried out, and experiments were verified through the system, the test results show that the system in the image of the high-altitude mountainous collection experiments , objects of different colors recognition rate of 100%, respectively, of the fire and terrain recognition rate 95.55% and 81.12%, very suitable for intelligent visual surveillance system under a variety of mountain environments.