基于DSP的全景视频多目标实时检测

全景视频由于具有360?的视角,非常适合用于智能视频监控中的多目标检测。首先介绍了以DSP处理器为核心的系统处理平台;接着介绍了全景图像的展开算法和MHOEI多目标检测算法;然后针对全景展开设计了适合DSP平台的分块查表展开方法,针对MHOEI设计了降分辨率处理方法;最后在DSP处理器上进行了优化实现。实验表明,在所选的DM6437处理器上,将720×576大小的全景图像展开为896×180大小的目标图像,展开帧率可达120帧每秒;再经过多目标检测后能达到平均35帧每秒,满足实时要求,同时能有效检测出全景视频中的运动和静止目标。     

基于镜面折反射全景成像的三维重建方法研究

针对基于图像的三维重建过程中成像系统获得的双视点或多视点图像不能一次性完整建模、图像匹配过程中计算复杂度高、标定繁琐的问题,设计了一种基于镜面折反射全景成像技术的单目多视角立体视觉系统及三维建模算法。该方法根据立体视觉传感器获取被测物体表面信息图像,通过阈值分割将图像分割为5个基本视图,并建立空间几何约束关系;使用离散的体素来表示被重构物体的几何信息,采用光一致性约束进行物体表面点的判断;通过轮廓体素遍历获取物体三维模型。该方法能一次性获得物体不同视角全方位图像信息,根据空间几何约束、纹理颜色约束及物体曲线约束,快速重构出物体的三维几何模型。该方法降低了匹配计算的复杂度和硬件使用成本,其有效性已在实际系统中得到了验证。     

全景图像实时校正系统的设计与实现

利用FPGA作为可见光全景图像处理器,使用硬件畸变校正系统,以达到对全景图像畸变场景的实时校正.全景图像畸变校正核心问题是像素空间位置校正,以CORDIC算法为基础,设计了全景环形图像展开算法,实现了全景环形图像的展开.设计采用ITU-656标准数字视频格式,用VHDL硬件描述语言实现整个校正算法的FPGA设计,试验结果表明,能实现全景环形图像的360.实时展开.     

柱面场景无畸变的折反射全景成像系统

提出了一种基于透视成像模型的柱面场景无畸变的折反射全景成像系统设计方法,该方法 利用系统柱面场景无畸变成像要求和视场角确定反射镜面形。根据这一方法设计制作了高次旋转对称非球面反射镜,并用该反射镜和CCD摄像头建立了全景成像系统,实现了对360°柱面场景的无畸变成像。     

谈720全景图片的拍摄与制作

正一、720全景图片的概念720全景图片是给人以三维立体感觉的720°全方位实景图像。它实际上是通过全景播放器的矫正处理而形成的三维全景,即水平360°、上下360°全能观看的720全景,也就是能看到包含"天与地"的全景。720全景拍摄是指固定一个位置点,进行水平360°和垂直360°的多次拍摄,并利用专门软件拼接所组成的球形全景,最终将所拍摄的内容拼接成一张长宽比为2:1的图片用于后期发布(如图1、图2)。720全景图为浏览者提供了各种操纵图像的功能,包括放大缩小、各个方向移动观看场景等,以达到模拟和再现场景的真实环境的效果,当720全景图和虚拟现实相结合,会产生全新的观赏体验,常用于房地产样板间实景展示、商业空间漫游及旅游景点实景展示等方面。     

基于前向映射的全景视觉图像解算方法

解算的主要目的是将全方位图像还原成无畸变的图像.本文根据双曲面折反射全景视觉系统成像特点,提出了一种基于前向映射的全方位图像解算的改进算法.该算法首先采用前向映射空间变换方法将全方位图像的像素映射到虚拟观察平面上,然后利用改进的Shepard散乱数据插值法复原完整的解算图像,从而避免了在畸变的全方位图像中直接插值引起的图像解算的失真.实验结果表明,该方法不但消除了全方位图像的非线性畸变,而且提高了插值精度和图像灰度曲面的光滑性,使解算图像更接近真实景象.     

基于Bezier曲面的快速插值算法改善全景图像的分辨率

全景环形透镜(PAL)是基于平面圆柱投影法将围绕光轴360°的视场投影到环形平面上,具有景深无限远,图像映射关系为线性特点,在机器人视觉、监控和虚拟现实领域有着重要的应用.实际应用中需要将环形像无失真的展开为常规平面像,此过程需要进行插值处理,三次样条插值精度较高,但是速度很慢.文章采用Bezier曲面的插值算法,并针对全景图像的特点,提出了Bezier曲面的全景图像快速插值算法,使得图像分辨率比样条插值和双线性插值得到改善的同时,插值处理时间比三次样条插值缩短了80%左右.     

野外监测折叠无人机

正设计说明:该无人机为大型六旋翼折叠多轴无人机,主要用于野外监测,森林草原防火监测,户外全景视频监控等领域,具有飞行载荷大、滞空时间长、可全景拍摄的特点。旋臂采用三幅折叠结构,每根支撑两个螺旋桨。折叠收缩后可紧密围绕在雷达天线的周围。载荷部分由三组可变焦摄像头实现了360°全景监控,并配备了红外热敏感应装置,     

基于光照-反射模型的图像增强算法的FPGA设计

针对视频图像增强处理的应用需求及视频图像场景的多样化特点,给出了一种基于光照-反射模型的低照度图像增强算法及其硬件实现。首先,基于光照-反射模型将亮度图像分解为照度分量、反射分量,并对估计的照度分量进行非线性调节,合成新的亮度图像;然后,引入调整系数对光照估计进行场景补偿以适应不同的场景;最后,在现场可编程逻辑门阵列(FPGA)硬件上进行算法实现。在Altera EP4CE40F23C8 FPGA硬件上实现结果表明,图像增强算法的主、客观处理效果得到明显提升,可满足不同场景的低照度视频图像实时增强处理应用。     

全景摄影:制造影像新感觉

什么是“全景摄影”或“全景照片”呢?全景摄影(Panorama)也叫作虚拟现实(Versual reality),是基于图像处理的虚拟漫游。全景摄影是把相机环360°拍摄的一组照片拼接成一幅全景图像, 通过对场景中360°全方位图像的摄取,程序的无缝拼接,以及视角变形,热点设定,给浏览者展现一个三维的空间。辞典中的定义称:其拍摄的视角应超过180°。1　全景摄影按拍摄方式分类　　全景摄影按拍摄方式可分为:柱型全景—垂直视角约60°,水平360°,适合表现高清晰的远景;球型全景—垂直180°,水平360°,即包括了完整的视域,拍摄时需要避开摄影架而多次定…     

Defocus deblurring for catadioptric omnidirectional imaging based on spatially invariant point spread function

The problem of catadioptric omnidirectional imaging defocus blur, which is caused by lens aperture and mirror curvature, becomes more severe when high resolution sensors and large apertures are applied. In order to overcome this problem, a novel method based on computational photography is proposed. Firstly, the defocus blur of catadioptric omnidirectional imaging is analyzed to calculate the point spread function for different scene points. Then, the defocus blur kernel of omnidirectional image is confirmed to be spatially invariant when rotating the focus ring of camera lens during an image’s integration time. Lastly, the deconvolution algorithm using prior sparse derivatives is applied to obtain all-focused/sharp omnidirectional images. Experimental results demonstrate that the proposed method is effective for omnidirectional image deblurring and can be applied to most existing catadioptric omnidirectional imaging systems.     

Incorporating lens distortion into the design of undistorted catadioptric omnidirectional cameras

An improved approach for design of the catadioptric omnidirectional camera with horizontal scene undistorted imaging is described. In the proposed approach, the influence of the lens distortion on the imaging quality of the omnidirectional camera is taken into account. We establish a radial distortion model for an image pickup lens and retain some opposite distortion in the mirror to correct the distortion existing in the image pickup lens. A horizontal scene undistorted catadioptric omnidirectional camera is designed with an off-the-shelf TV short focus lens using our approach; the numerical simulation shows that the distortion introduced by the imaging lens is eliminated effectively.     

一次成像折反射式步进变焦距物镜

设计了一个用于远距离监视电视的折反射式步进变焦距透镜。为使微光、白光和近红外三个波段共用一个光学系统,提出一次成像折反射式变焦,并进行了理论分析和给出了设计实例。克服了二次成像折反射式步进变焦方式二级光谱不能校正的缺陷,并且杜绝了像面中心产生黑斑的可能,结构也比较简单。     

An ultrashort throw ratio projection lens design based on a catadioptric structure

In this paper, we present a rotational symmetry for an ultrashort throw (UST) lens with offset field. The UST lens has a throw ratio of 0.23 and a total track of 195 mm. The optical elements of the UST lens are comprised of two parts. First, a catadioptric projection lens where the catadioptric function permits reaching an ultrashort throw ratio, short total track, while at the same time requiring fewer lens elements. The second part is a collimating lens which takes advantage of the telecentric condition to generate uniform total internal reflection (TIR) in the TIR prism. With this design, an effective focal length of -1.96 mm and a f-number of 2.4 can be obtained. The root mean square spot size and lateral colour of all fields are smaller than one pixel in size. The maximum optical distortion of ?0.97% and TV distortion of 0.2% are acceptable. In terms of image quality, the modulation transfer function (MTF) values for all fields are above 0.65 at 0.245 line pairs/mm. Even when the tolerance error is considered, the MTF values for all fields are still above 0.3. The suitability of the novel UST lens design for projection applications is discussed.     

Houghton anastigmatic telescope containing aspheric mirrors that allows corrector position as a free parameter

A family of catadioptric telescopes is investigated, characterzied with two-element, full-aperture afocal corrector lenses and aspherical, focusing, primary mirrors or primary-secondary mirror combination. Third-order aberration and design equation are provided for anastigmatic systems in which corrector position must be a free parameter, forcing the consideration of aspherics on the mirrored surfaces.     

Optical Systems for Use at Low Light Levels

The development of image intensifier tubes and their application in instruments for operation under low light-level conditions created the need for optics which require specific properties. Recognition of the problems involved when combining a lens with a tube results in general requirements for the lens, some of which are different from those in standard lens design. Refined basic lens configurations and new systems have been introduced in this field, e.g. refractive and catadioptric lenses. Some examples of their performance are shown. In addition relay lenses having an extremely high speed have been realized for applications mainly for x-ray image intensifier systems. Single-tube instruments with a bi-ocular eyepiece are shown to have many advantages over the double-tube binocular.     

折反射式连续变倍扩束系统的设计

为了用同一台仪器获得不同口径的激光准直光束,设计了一个连续变倍激光扩束系统,系统扩束比可在55～155之间连续变化,可以得到比传统扩束系统更大的激光光斑,满足对大口径激光束应用的要求.该系统由两级扩束子系统构成,一级折射式无焦变倍扩束系统实现激光光束口径连续变化,变焦系统选用仅有4片透镜的三组元结构,并采用"正-负-正"的形式,有利于减小系统整体结构尺寸.考虑到可能产生的热形变,对入射光斑较小的变倍移动组的材料进行了分析选择.二级折反射系统实现激光束的高倍率扩束,由球面反射镜取代离轴抛物面镜,用一片弯月透镜补偿像差.由于单透镜折射元件对系统性能影响很大,对其结构形式和材料进行了分析,确定了透镜的厚度与材料.应用光学设计软件CODEV给出了系统的优化设计,设计结果显示该系统像差得到了很好地校正,成像质量达到衍射极限.最后用Light tools软件进行三维建模分析,系统准直度与理论设计值相符.     

Folded catadioptric panoramic lens with an equidistance projection scheme

A new formula for a catadioptric panoramic lens with an equidistance projection scheme has been derived. The fabricated lens has a field of view that is wider than that of any previously reported panoramic lens, and the nonimaged region near the back of the camera has a constant volume with zero angular extension.     

A triangle mesh-based corner detection algorithm for catadioptric images

As with conventional images corner detection is an important aspect of many computer vision problems involving catadioptric images. However, classical image processing algorithms are no longer appropriate for catadioptric images due to nonuniform resolution and distortions of catadioptric images. In this paper, we propose a novel approach to corner detection for catadioptric images based on triangle mesh. First, we transform catadioptric images to spherical images by combining an improved projection model for central catadioptric cameras with triangle mesh for a unit sphere. Spherical images yield a spatially uniform resolution domain for processing catadioptric images. Then, based on the topology of a triangle mesh, variations of light intensity with respect to directions for each image patch are measured to detect corners. The proposed algorithm addresses problems of catadioptric image processing caused by non-uniform resolution and distortions of these images. Experimental results showed that comparing to widely used methods, the triangle mesh-based corner detection algorithm can achieve higher repeatability rate relative to different imaging condition changes.     

Wide-angle catadioptric lens with a rectilinear projection scheme

A catadioptric wide-angle lens having a rectilinear projection scheme has been developed with a view to possible applications in the security-surveillance area. The lens has been designed for a miniature camera with a video graphics array-grade 1/3 in. color CCD sensor. The field of view of the lens is over 151 degrees , and still distortion is under 1%. Furthermore, the modulation transfer function is better than 0.3 at 70 line pairs/mm over the whole active area of the image sensor.