[论文格式样板.docVIP

摘 要 本论文主要研究油罐液位监测系统中现场采集红外图片的处理，以提取油罐中油层与空气层，油层与沙层分界线在油罐的具体位置，为油罐液位测量，油罐综合管理提供便利。 论文提出一种新型的油罐液位测量方案，由安装在云台的上红外摄像头对油库区油罐进行拍照，在PC上对现场采集的图片进行预处理、边缘检测、横向液位特征提取最后通过文本文件输出，将图片分析结果存储在硬盘上，方便液位信息的实时显示。 边缘检测算法是图像处理的基础，经历长久发展目前已经现对成熟。论文结合油罐红外成像特点，提出了一种基于Canny准则的改进型自适用边缘检测算子。改进型自适用双阈值边缘检测算子，能够根据图片灰度自动设置高低阈值，避免了传统Canny检测算子需要人工设置高、低阈值的缺点，是检测算法更具有普遍性。 关键词：液位监测，油罐，边缘检测，红外 Abstract This paper studies tank level monitoring system in the infrared images collected at the scene of processed to extract oil in the reservoir and the air layer, oil and sand dividing line between the specific location of the tank for tank level measurement, the oil tank To facilitate the integrated management. Paper presents a new type of tank level measurement programmer, installed in Haeundae by the infrared camera on the tank depot area to take pictures on the PC on-site collection of pictures pretreatment, edge detection, horizontal solution Feature extraction by the end of the text file output, the picture of the results stored on the hard drive to facilitate the level of real-time display of information. Edge detection algorithm is the basis of image processing, has experienced long-term development of mature now. Papers tank with infrared imaging features, a Canny criteria based on the application of improved since the edge detection operator. Since the application of improved edge detection operator, according to Gray picture automatically set the threshold level to avoid the traditional Canny operator need to set up artificial high and low threshold of the shortcomings of detection algorithm is more universal. Key words:　Level monitoring, tank of oil, edge detection, infrared 目 录 摘 要 I Abstract II 1. 绪论 1 1.1 课题来源、目的和意义 1 1.2 国内外油罐监测研究现状 1 1.3 主要研究内容 2 1.4 本文结构 3 2. 油罐红外成像特点 4 2.1 油罐温度分布规律 4 2.2 红外热像仪成像原理 5 2.3 图像基本知识 5 2.3.1 常见图像文件格式 5 2.3.2 单色图像 6 2.3.3 灰度图像 7 2.3.4 RGB色彩模式 8 2.4 油罐图片目视分析 9 2.5 油罐图片灰度值分析 10 2.6 本章小结 10 3. 图像预处理 12 3.1 滤波除噪 12 3.1.1 邻域平均法 13 3.1.2 中值滤波 14 3.1.3 形态学除噪 1