嵌入式气体检测系统及其图像分析算法  

Embedded gas detection system and its image analysis algorithm

作　　者：罗小刚[1] 汪德暖[1] 柏兴洪[1] 

机构地区：[1]重庆大学生物工程学院,重庆400030

出　　处：《计算机应用》2011年第8期2270-2274,2278,共6页journal of Computer Applications

基　　金：国家863计划项目(2006AA04Z349);重庆市科委重点攻关项目(2008AB2024);重庆市科委自然科学基金资助项目(2009BB5219);"211工程"三期建设资助项目(S-09104);中央高校基本科研业务费资助项目(CDJXS102300)

摘　　要：针对传统气体检测方法的不足,提出了基于嵌入式和卟啉阵列传感器的气体快速检测系统及其图像分析算法。该系统通过USB接口的摄像头采集待测气体与卟啉阵列传感器反应前后的RGB图像信息,再由图像处理算法得到图像中每个卟啉点的颜色变化信息,最后通过模式识别算法得到待测气体种类和浓度。重点研究了检测系统的结构及软件功能的设计,给出了针对卟啉阵列传感图像的图像处理算法及与标准数据库匹配的模式识别算法。通过对氨气等气体进行试验,结果表明该系统及其分析算法能够很好地识别出气体的种类和浓度。A fast gas detection system and its image analysis algorithm based on embedded platform and porphyrin sensor arrays were designed to overcome the deficiency of the traditional gas detection methods.The system captured the images before and after porphyrin sensor arrays contacted with gases using the camera with USB interface at first.Then the color change information of each porphyrin spot in the images could be obtained by using the image processing algorithm.Finally the type and the concentration of unknown gas were recognized using pattern recognition algorithm.In addition,the design of the system structure and its software function were proposed.The image processing algorithm for porphyrin sensor array images and pattern recognition for matching the records in standard database were given.Lots of tests had been done for Ammonia and other gases.The experimental results show that the system and analysis algorithm can identify the type and concentration of gas.

关 键 词：气体检测 嵌入式系统 图像处理 模式识别 

分 类 号：TP212.2[自动化与计算机技术—检测技术与自动化装置] TP391.413[自动化与计算机技术—控制科学与工程]