基于Zynq的智能相机前端系统的研究
发布时间:2019-05-13 11:05
【摘要】:智能相机是机器视觉领域的研究热点之一,它集图像采集、处理、分析、传输等功能于一体,小型化、分布式、网络化等特点使其在不同行业有着广泛的应用前景。论文重点研究了基于Zynq平台智能相机的前端图像采集系统,探讨了CCD片内数据流传输模型、图像采集系统电路、CCD驱动逻辑IP核、图像采集模式、CCD性能参数评估等关键技术及其实现方法,主要工作包括: 在深入分析Zynq平台特性的基础上规划了基于该平台的智能相机的总体架构,完成图像采集、核心处理控制、以太网通信、显示、外部存储等主要功能模块的划分,建立了CCD片内数据流传输模,并采用XSVI总线作为通用内部数据传输接口。 针对ICX618图像传感器,设计了相应的CCD驱动外围、AD采样转换、CCD驱动电源等核心电路模块,采用CXD1267AN进行CCD三级驱动电平转换,,AD9949进行图像信号数模转换和部分控制信号驱动。以实现驱动和采集模式的通用多功能性为目标,采用有限状态机描述图像采集驱动的复杂逻辑,通过Verilog硬件描述语言实现具有多模块复杂协同特点的IP核开发,主要功能模块包括:CCD曝光、电荷读出、转移、检测等不同阶段的驱动时序;AD寄存器SPI配置协议;XSVI通用内部数据传输接口;彩色Bayer阵列双线性插值恢复逻辑。 完成智能相机图像采集前端系统样机开发工作,可实现内/外触发、单帧/连续采集、彩色/黑白成像等多种图像采集模式。对系统的驱动信号进行了完整性分析,通过暗电流噪声、响应度、动态范围等测试进行全面的CCD性能参数评估,并针对光学圆标记点检测进行初步应用测试。试验结果表明,采集系统图像性能达到预期要求,系统架构和流程实现方式满足智能相机功能需求。
[Abstract]:Intelligent camera is one of the research hotspots in the field of machine vision. It integrates image acquisition, processing, analysis, transmission and other functions. Miniaturization, distributed, networking and other characteristics make it have a wide range of application prospects in different industries. This paper focuses on the front-end image acquisition system based on Zynq platform intelligent camera, and discusses the on-chip data flow transmission model of CCD, the circuit of image acquisition system, the IP core of CCD driving logic, and the image acquisition mode. The key technologies such as CCD performance parameter evaluation and their implementation methods include: based on the in-depth analysis of the characteristics of the Zynq platform, the overall architecture of the intelligent camera based on the platform is planned, and the image acquisition and core processing control are completed. By dividing the main functional modules such as Ethernet communication, display, external storage and so on, the data stream transmission model on CCD chip is established, and XSVI bus is used as the general internal data transmission interface. For ICX618 image sensor, the corresponding CCD drive peripheral, AD sampling conversion, CCD drive power supply and other core circuit modules are designed. CXD1267AN is used for CCD three-level drive level conversion, AD9949 for image signal digital-to-analog conversion and partial control signal drive. In order to realize the general versatility of driving and acquisition mode, the complex logic of image acquisition driver is described by finite state machine, and the IP core development with the characteristics of multi-module complex cooperation is realized by Verilog hardware description language. The main functional modules include: CCD exposure, charge readout, transfer, detection and other different stages of the driving timing; AD register SPI configuration protocol; XSVI general internal data transmission interface; color Bayer array bilinear interpolation recovery logic. The prototype of the front-end system of intelligent camera image acquisition can be developed, which can realize many image acquisition modes, such as internal / external trigger, single frame / continuous acquisition, color / black-and-white imaging and so on. The integrity of the driving signal of the system is analyzed, the performance parameters of CCD are evaluated by dark current noise, responsiveness, dynamic range and so on, and the preliminary application test is carried out for the detection of optical circular marking point. The experimental results show that the image performance of the acquisition system meets the expected requirements, and the system architecture and process implementation meet the functional requirements of the intelligent camera.
【学位授予单位】:华南理工大学
【学位级别】:硕士
【学位授予年份】:2014
【分类号】:TB852.1
本文编号:2475827
[Abstract]:Intelligent camera is one of the research hotspots in the field of machine vision. It integrates image acquisition, processing, analysis, transmission and other functions. Miniaturization, distributed, networking and other characteristics make it have a wide range of application prospects in different industries. This paper focuses on the front-end image acquisition system based on Zynq platform intelligent camera, and discusses the on-chip data flow transmission model of CCD, the circuit of image acquisition system, the IP core of CCD driving logic, and the image acquisition mode. The key technologies such as CCD performance parameter evaluation and their implementation methods include: based on the in-depth analysis of the characteristics of the Zynq platform, the overall architecture of the intelligent camera based on the platform is planned, and the image acquisition and core processing control are completed. By dividing the main functional modules such as Ethernet communication, display, external storage and so on, the data stream transmission model on CCD chip is established, and XSVI bus is used as the general internal data transmission interface. For ICX618 image sensor, the corresponding CCD drive peripheral, AD sampling conversion, CCD drive power supply and other core circuit modules are designed. CXD1267AN is used for CCD three-level drive level conversion, AD9949 for image signal digital-to-analog conversion and partial control signal drive. In order to realize the general versatility of driving and acquisition mode, the complex logic of image acquisition driver is described by finite state machine, and the IP core development with the characteristics of multi-module complex cooperation is realized by Verilog hardware description language. The main functional modules include: CCD exposure, charge readout, transfer, detection and other different stages of the driving timing; AD register SPI configuration protocol; XSVI general internal data transmission interface; color Bayer array bilinear interpolation recovery logic. The prototype of the front-end system of intelligent camera image acquisition can be developed, which can realize many image acquisition modes, such as internal / external trigger, single frame / continuous acquisition, color / black-and-white imaging and so on. The integrity of the driving signal of the system is analyzed, the performance parameters of CCD are evaluated by dark current noise, responsiveness, dynamic range and so on, and the preliminary application test is carried out for the detection of optical circular marking point. The experimental results show that the image performance of the acquisition system meets the expected requirements, and the system architecture and process implementation meet the functional requirements of the intelligent camera.
【学位授予单位】:华南理工大学
【学位级别】:硕士
【学位授予年份】:2014
【分类号】:TB852.1
【参考文献】
相关期刊论文 前10条
1 李相民,周立伟,金伟其;科学级致冷 CCD 相机的性能和技术[J];大连理工大学学报;1997年S2期
2 刘焕军,王耀南,段峰;机器视觉中的图像采集技术[J];电脑与信息技术;2003年01期
3 马天波;刘辉;臧佳;;线阵CCD性能指标测试软件设计[J];光机电信息;2011年07期
4 冉峰;杨辉;黄舒平;;面阵CCD彩色视频图像实时采集系统的设计[J];光学精密工程;2010年01期
5 朱广;周程亮;;CCD图像传感器在路径识别系统中的应用[J];工业控制计算机;2009年03期
6 许秀贞,李自田,薛利军;CCD噪声分析及处理技术[J];红外与激光工程;2004年04期
7 程耀瑜,胡潬,韩焱;化妆品级制冷CCD的研究及其在X射线成像检测中的应用[J];激光与光电子学进展;2002年02期
8 侯新梅;李自田;胡炳j;;AD9822及其在面阵CCD系统中的应用[J];现代电子技术;2010年20期
9 张林;李永新;胡学友;;基于相关双采样技术的CCD视频信号处理研究[J];宇航计测技术;2007年02期
10 姜忠宝;高俊国;段少丽;王昌盛;;CCD图像传感器的原理及军事应用[J];电子元器件应用;2003年01期
相关博士学位论文 前1条
1 冯晓静;面向服务的异构多核片上系统的关键技术研究及实现[D];中国科学技术大学;2013年
本文编号:2475827
本文链接:https://www.wllwen.com/guanlilunwen/gongchengguanli/2475827.html