新型高速视频图像记录判读系统

介绍了新型高速视频图像记录判读系统的特点、组成和功能,并采用了一种改进的自动判读方法.该系统利用集成在高速CMOS摄像机上的存储器记录数字视频序列图像,并利用高速接口下载到计算机移动硬盘或可读写光盘上,以供事后使用判读软件对记录的图像进行分析处理.运用数学形态学图像处理方法滤除二值图像的噪声,较好地实现了清晰提取目标边界的目的.针对自动判读与半自动判读的不同要求,采用不同的细分技术.在现有分辨率的情况下,达到对目标边缘亚像元级定位的目的,显著提高测量准确度.对其他类似系统的研制有一定的借鉴作用.     

视频判读系统分析

本文就光电测量提出了以CCD进行视频捕获,将获得的视频图象通过采集卡滤波、A/D转换为数字图象,存储在计算机内存中,根据具体要求选择相应的算法,由计算机对其进行分析处理.     

高速红外图像判读系统设计研究

针对实时跟踪领域事后图像判读处理数据量大、时间冗长的问题，提出了一种新型高速图像判读系统．该系统基于多线程乒乓缓冲技术，实现了图像加载、显示及处理三项工作相互独立和并行执行．通过对红外图像进行灰度拉伸，预处理，亚像元插值细分等操作，有效提高了判读准确度．实验表明，该判读系统提高了数字图像判读系统的工作效率，大大缩短了数据处理时向．     

基于双PCI总线的高速视频记录方法

基于目前图像传感器和计算机软硬件技术水平，提出了一种实现超高速视频流数据长时间连续记录方法。以DSP为核心的控制电路直接驱动图像传感器的视频流，将其一分为二，并转换为两路标准数字视频信号，经由两路高路PCI总线记录到硬盘。同时详细讨论了性能优异的卡对卡技术和适合随机存取的RAID技术实现连续记录的方法。     

视频图象分析处理系统应用研究

本文描述了一种基于视频图象的分析处理系统,结合工程应用实践,提出利用微机终端显示器作为象框坐标测量器,实现对视频图象的分析处理.不仅具有丰富的数字图象处理功能,还可对原始图象进行增强和细分等处理.     

高分辨率航拍图像实时记录方法

针对某型高分辨率航空摄像机输出数据流达到每秒200兆字节,常规方法难于实现图像数据实时记录的问题,提出一种自启动多通道混合同步直写SCSI硬盘的方法.通过新设计的状态控制器,优化了SCSI硬盘的记录时序,使多个并行记录通道能根据数据缓存器和硬盘的状态自动产生独立的直写脉冲,以混合同步的方式将数据高速记录到硬盘上.实验结果表明:该方法可使图像数据持续记录速率达到每秒200兆字节,储存容量可达280吉字节,且结构简单、实用.     

高分辨率航拍图像实时记录方法

针对某型高分辨率航空摄像机输出数据流达到每秒200兆字节,常规方法难于实现图像数据实时记录的问题,提出一种自启动多通道混合同步直写SCSI硬盘的方法.通过新设计的状态控制器,优化了SCSI硬盘的记录时序,使多个并行记录通道能根据数据缓存器和硬盘的状态自动产生独立的直写脉冲,以混合同步的方式将数据高速记录到硬盘上.实验结果表明:该方法可使图像数据持续记录速率达到每秒200兆字节,储存容量可达280吉字节,且结构简单、实用.     

视频内窥镜图像采集与处理

提出一种用ＣＣＤ接收图像的视频内窥镜系统，其图像用微机进行处理，实验结果表明由该系统获得的人耳图像经微机处理后，图像的清晰度有明显改善。     

高速视频硬盘记录系统分析

为了提高视频记录速度,通过分析视频记录系统的数据通道,找出通道的瓶颈所在;通过研究文件大小与存储速度的关系,采用RAID 0技术和文件合成与分割技术提高硬盘的存储速度,改善了系统的数据通道,最终达到了提高视频记录系统的整体速度,保证了记录数据的实时性和完整性.     

基于光学图像的目标姿态判读处理

 采用摄像测量方法获取目标的飞行姿态是目前靶场主要测量手段之一。目标姿态可以用目标轴线在空间的方位来描述,即姿态的判读处理就是要提取线的空间位置。在进行图像判读处理时除采用传统的基于点的判读处理手段外,还应根据不同的目标类型采用不同的判读处理方法,文中给出了基于首尾两点的判读处理、基于线的判读处理和基于模型的三种判读处理方法,有效解决了不同类型目标的判读处理问题。     

实时图像记录与处理系统

针对图像记录与处理的实时性问题,设计并实现了实时图像记录与处理系统,同时实现了对高速图像采集、海量数据存储和实时图像处理功能。介绍了系统的硬件组成和图像实时记录、处理的软件设计,在实时图像处理中将最小二乘法应用于角点检测,实验结果表明该方法检测精度能达到亚像素级,而且对噪声不敏感,检测稳定性好。     

Shingle magnetic recording assessment with spinstand measurement

Shingle Magnetic Recording (SMR) is a candidate to realize recording density of over 1 Tbit/in². At the cost of complex HDD firmware, we can expect many gains such as higher write-ability to break through the famous tri-lemma. In this paper, promising empirical data and concerns found on spinstand were introduced and discussed. More than 18 dB of better reverse overwrite (ROW) than conventional write recording (CWR) was confirmed at 40 nm of track pitch. Areal density capability (ADC) was improved by 25.5% by SMR. Another gain from SMR was observed at inner diameter (ID) and outer diameter (OD) that ADC loss due to skew effect was improved by using preferable side write pole edge. The dependence of ADC on magnetic read width measured with micro-track method (MRWu) indicates that side reading is limiting ADC. Reduction in side reading through narrower MRWu and inter-track interference cancellation technology are necessary to further improve ADC. BPI capability improvement by steeper down track field gradient at the track edge needs to be considered.     

A hardware-software system for analysis of video images

The purpose of this paper is to describe a software/hardware systemfor the analysis of digitized video images and a number of applications for which it may be used. The system described includes a Macintosh computer, a frame-grabber board, and Image, a public domain software program available at no cost from the U.S. National Institutes of Health. In our clinic and laboratory, this system is routinely used to make quantitative measurements from videofluoroscopic x-ray images of dynamic swallow studies and studies performed to assess velopharyngeal dysfunction in speech. It can also be used to examine various laryngeal parameters obtained from videotaped endoscopic and stroboscopic examinations. With a videocamera attached to a microscope, the system permits quantitative analysis of tissue characteristics, e.g., thickness of epithelial or connective tissue layers of the vocal folds. The relatively low cost and ease of use of the image analysis system make it a particularly attractive option when quantitative assessment of clinical or research materials in video format is desirable.     

Extensions of perpendicular recording

Gains in storage density in magnetic recording have fundamentally been achieved by scaling—reduced geometrical dimensions under the assumption that the recording physics does not change if all dimensions are scaled appropriately. It is becoming clear that evidence of the breakdown of scaling is now seen. We will here discuss ways to break the constraints on magnetic recording set by scaling. In particular, we will discuss energy-assisted recording, domain-wall-assisted recording, and bit-patterned media, with some emphasis on recording system considerations.     

Exploitation of inter-track interference in a shingled replay system

The expected increase in areal density in hard drives will require very narrow tracks. Tracks which are of a similar width to the read head and which are not separated by guardbands normally suffer from large inter-track interference (ITI) or crosstalk. Here, we show that it is possible to read from tracks which are not separated by guardbands and that are narrower than the head. In addition to the significant increase in areal density obtainable by reducing the unrecorded area of the disk and narrowing the tracks, such a system would also lead to a decrease in data retrieval times. We have identified across-track magnetization constraints for future coding across three adjacent tracks so that it will be possible to read from tracks which are only 73% as wide as the read head. Reading from tracks not separated by guardbands which have been written under these constraints yields an increase in track density of at least 47% greater than that possible in conventional drives.     

多阶光存储原理分析及其读写系统前端信号处理研究

运用矢量衍射理论的计算方法(耦合波方法，时域有限差分法)，对两种多阶光存储方案(坑深调制，坑形调制)进行了原理性分析，并简略分析了利用相变材料的幅值调制实现多阶存储的原理.由此得出多阶光存储技术的实现是利用了光与物质之间的相互作用，使光对记录介质产生作用，同时介质也对光的某些特性进行调制，而这些特征的改变一般都不止两个状态0或1，用数字方式将这多个状态记录下来便实现了多阶数字存储.此外,还提出了多阶光盘前端信号处理系统的设计与开发.该系统是多阶光存储实验系统的重要组成部分，可输出RF,TE,FE等伺服与误差信号，并将信号反馈给多阶处理芯片；同时系统还可对光学头的激光读写功率及写入时间进行控制，并能使其实现自动调焦和道跟踪. 关键词： 多阶光存储 矢量衍射理论 时域有限差分法     

弹道终端飞机目标红外图像瞄准点识别方法

依据飞机目标红外图像的特点,将飞机机头选定为跟踪瞄准点。利用弹道终端目标未充满视场前的红外序列图像、视场中心角、弹目距离等,建立了弹体坐标系下飞机轴线方向的求解模型。采用MAT、Hough变换等技术,对目标充满视场后的红外序列图像进行处理,检测出了飞机轴线,识别出了飞机机头。利用从弹道终端的两个序列图像中抓取出来的四帧红外图像进行了仿真,并给出了实验结果。     

Below-diffraction-limited hybrid recording using silicon thin film super-resolution structure

We report on new experimental results for below-diffraction-limited hybrid recording. In our experiments, by means of focused laser assisted magnetic recording, the magnetic domains within TbFeCo thin films are obtained under an external perpendicular direct magnetic field. For a single magnetic medium, the domain size is mainly determined by the focused spot, which is about 620~nm for the laser wavelength λ =406~nm, and a numerical aperture of the lens of 0.80. However, when a silicon thin film structure is inserted between the substrate and the magnetic medium, the recording domains can be reduced obviously. By optimizing the experimental condition, even the size can be reduced to about 100~nm, which is below the diffraction limit, i.e. about 1/6 of the spot size. This is very useful for improving the hybrid recording density in practical applications.     

新型防潮的光致聚合物全息记录材料及其应用

本文研制出一种新型兰敏的光致聚合物全息记录材料,并提出再聚会固定全息再现峰值波长的方法.这种记录材料由增感剂,引发剂,链转移剂,单体和成膜物组成.用本材料记录的全息图衍射效率可达95%以上,且能防潮,耐高温.