基于红外热成像的车载夜视系统

发布时间：2018-05-29 02:45

本文选题：红外热成像 + 夜视系统　；参考：《华南理工大学》2015年硕士论文

【摘要】：基于红外热成像的车载夜视系统是一种利用红外热成像技术,做为车辆的主动安全设备,辅助驾驶员在夜间及能见度低的情况下看清前方路况,并对车辆前方的行人进行识别,协助驾驶员对前方路况及行人所处的位置做出预判,提前预防危险,减少交通事故的发生。论文首先介绍车载红外夜视系统对汽车夜间驾驶的重要作用,及国内外车载红外夜视系统的研究现状；然后对红外线的发现、红外热成像原理、红外热像仪组成做简单的介绍；接下来又介绍了行人识别的基本原理,包含图像分割、模板匹配、机器学习等；最后根据车载红外夜视系统的要求,给出系统的总体设计及各功能模块的实现。车载红外夜视系统主要包含车载红外热像仪和行人识别模块两大部分。车载红外热像仪设计部分主要包含车载红外热像仪的组成、红外镜头的选型、红外探测器的选型、校正组件设计、窗口加热模块实现、电源电路设计、红外探测器驱动电路设计、红外探测器控温电路设计、红外采样电路设计、控制及图像处理电路设计、视频输出电路设计等。行人识别系统设计主要包含系统组成、硬件设计、行人识别算法介绍等。该系统红外探测器采用法国ULIS公司生产的UL03262型非制冷焦平面红外探测器,分辨率为384*288,感应红外波段为8-14μM,红外探测器接收到目标的红外辐射后经光电转换输出带有图像信息的模拟信号,用高速ADC对该信号进行采集,输送给FPGA,FPGA与单片机相互配合对红外图像进行降噪、盲元替换、非均匀性校正、灰度变换、字符叠加等操作,输出给外部视频解码器,最终转化成复合视频信号(CVBS)传送给行人检测模块,经图像分割、模块匹配、行人判别等操作,判定图像中是否有行人,并将判定信息叠加后输送到车载影音系统,进行显示。
[Abstract]:The vehicle night vision system based on infrared thermal imaging is an active safety device for vehicles using infrared thermal imaging technology to assist drivers to see clearly the road conditions in front of vehicles at night and under low visibility, and to identify pedestrians in front of vehicles. Assist the driver to prejudge the road condition and the position of the pedestrian in advance to prevent the danger and reduce the traffic accident. Firstly, the paper introduces the important role of infrared night vision system in vehicle driving at night, and the research status of infrared night vision system at home and abroad, and then introduces the discovery of infrared ray, the principle of infrared thermal imaging and the composition of infrared thermal imager. Then it introduces the basic principle of pedestrian recognition, including image segmentation, template matching, machine learning, etc. Finally, according to the requirements of infrared night vision system, the overall design of the system and the implementation of each functional module are given. Vehicle infrared night vision system mainly includes two parts: vehicle infrared thermal imager and pedestrian identification module. The design part of the vehicle infrared thermal imager mainly includes the composition of the vehicle infrared thermal imager, the selection of infrared lens, the selection of infrared detector, the design of calibration module, the realization of window heating module, the design of power supply circuit. Infrared detector drive circuit design, infrared detector temperature control circuit design, infrared sampling circuit design, control and image processing circuit design, video output circuit design. Pedestrian recognition system design mainly includes system composition, hardware design, pedestrian recognition algorithm introduction and so on. The infrared detector of this system adopts UL03262 uncooled focal plane infrared detector, which is produced by ULIS Company of France. The resolution is 384C 288, the infrared band is 8-14 渭 M. the infrared detector receives the infrared radiation from the target and outputs the analog signal with image information by photoelectric conversion. The signal is collected by high speed ADC. It is transmitted to FPGA and MCU to perform noise reduction, blind element replacement, nonuniformity correction, gray conversion, character superposition and so on, and output to the external video decoder. Finally, the composite video signal CVBS is transferred to the pedestrian detection module, which is used for image segmentation, module matching, pedestrian discrimination, and so on, to determine whether there are pedestrians in the image, and to transfer the decision information to the vehicle audio and video system after superposition for display.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U463.6;TN215

【相似文献】