几种深度摄像头简介

摘自：http://www.cvchina.info/2010/06/28/brief-review-of-several-depth-camera/

随着微软kinect(natal)的发展，深度摄像头吸引越来越多人的目光，深度摄像头可以用在人体跟踪，三维重建，人机交互，SLAM等等领域。但是深度摄像头的高昂的价格实在是让一般人望而却步，我所知道的primesense
的一个摄像头要5000美元。。。而kinect的出现会不会带动民用（相对廉价）的深度摄像头的发展呢（传闻kinect定价199美元）？
cvchina曾经介绍过两种深度摄像头，下面再转一个最近看到的几种深度摄像头的简介：

The Mesa Imaging SwissRanger 4000 (SR4000) is probably the most well-known ToF depth camera.  It has a range of 5-8 meters, 176 x 144 pixel resolution over 43.6° x 34.6° field of view.  It operates at up to 54 fps, and costs about $9,000.  I’ve seen these used in a number of academic laboratories.The PMD Technologies CamCube 2.0 is a lesser-known, but equally impressive ToF depth camera.  It has a range of 7 meters, 204 x 204 pixel resolution with 40.0° x 40.0° field of view.  It operates at 25 fps, and last time I checked, it costs around $12,000.

As part of the PR2 robot, Willow Garagehas developed a dense stereo rig using visible projected light (embedded in the sensor head pictured above-left).  I refer to this class of depth camera system as “projected-light stereo.”  In essence, a projector displays (pseudo) random patterns onto the scene to enable robust stereo feature extraction.  This method seems to produce very high-quality range data at framerate to produce point-clouds like the one shown below-left.  The Willow Garage system was (partially) explored in a paperat Humanoids 2009 by Rusu et. al.The biggest drawback with the system in its current form is that it uses visible light.  I’m sure subsequent versions will transition to infrared.

To my knowledge, these are not yet available for purchase.

The sensor pictured is Microsoft’s Project Natal, as announced at 2009′s E3 convention.  A brief history…  Microsoft purchased ZCamfrom 3DV Systems in 2009.  ZCam had made some impressive ToF imagers in the past and had plans to make a custom-silicon solution for ranging webcams.  However, the safe bet is that Project Natal is actually infrared projected-light stereo, hence the three opening ports on the pictured device (two camera ports for stereo imaging, one extra port for IR projection). Presumably this device will be quite cheap; it is meant to be a mass-market gaming console sensor.

To my knowledge, these are not yet available for purchase.

The PrimeSense depth camera is driven by a single-chip custom-silicon solution.  It projects a known infrared pattern onto the scene (perhaps a grid-like pattern?) and determines depth based on the pattern’s deformation as captured by an infrared CMOS imager.  The resulting depth image is 640 x 480 pixels with a maximum throughput of 60fps.  An optional color CMOS imager returns 1600 x 1200 photos.  You can learn more here.

To my knowledge, these are not yet available for purchase.

TOF是Time of flight的简写，直译为飞行时间的意思。所谓飞行时间法3D成像，是通过给目标连续发送光脉冲，然后用传感器接收从物体返回的光，通过探测光脉冲的飞行（往返）时间来得到目标物距离。这种技术跟3D激光传感器原理基本类似，只不过3D激光传感器是逐点扫描，而TOF相机则是同时得到整幅图像的深度信息。TOF相机与普通机器视觉成像过程也有类似之处，都是由光源、光学部件、传感器、控制电路以及处理电路等几部单元组成。

照射单元
TOF技术采用主动光探测方式，与一般光照需求不一样的是，TOF照射单元的目的不是照明，而是利用入射光信号与反射光信号的变化来进行距离测量，所以，TOF的照射单元都是对光进行高频调制之后再进行发射，比如下图所示的采用LED或激光二极管发射的脉冲光，脉冲可达到100MHz。

光学单元
与普通相机类似，TOF相机芯片前端需要一个搜集光线的镜头。不过与普通光学镜头不同的是这里需要加一个带通滤光片来保证只有与照明光源波长相同的光才能进入。同时由于光学成像系统具有透视效果，不同距离的场景为各个不同直径的同心球面，而非平行平面，所以在实际使用时，需要后续处理单元对这个误差进行校正。

TOF芯片
作为TOF的相机的核心，TOF芯片每一个像元对入射光往返相机与物体之间的相位分别进行纪录。该传感器结构与普通图像传感器类似，但比图像传感器更复杂，它包含2个或者更多快门，用来在不同时间采样反射光线。因为这种原因，TOF芯片像素比一般图像传感器像素尺寸要大得多，一般100um左右。

驱动电路：
照射单元和TOF传感器都需要高速信号控制，这样才能达到高的深度测量精度。比如，照射光与TOF传感器之间同步信号发生10ps的偏移，就相当于1.5mm的位移。而当前的CPU 可到3GHz，相应得时钟周期是300ps，则相应得深度分辨率为45mm。

运算单元
运算单元主要是完成数据校正和计算工作，通过计算入射光与反射光相对相移关系，即可求取距离信息。

TOF的优势：

系统简单
与立体相机或三角测量系统比，TOF相机体积小巧，跟一般相机大小相去无几，非常适合于一些需要轻便、小体积相机的场合。
高速处理
TOF相机能够实时快速的计算深度信息，达到几十到100fps。TOF的深度信息。而双目立体相机需要用到复杂的相关性算法，处理速度较慢。
高效运算
TOF的深度计算不受物体表面灰度和特征影响，可以非常准确的进行三维探测。而双目立体相机则需要目标具有良好的特征变化，否则会无法进行深度计算。
精度稳定
TOF的深度计算精度不随距离改变而变化，基本能稳定在cm级，这对于一些大范围运动的应用场合非常有意义。

摘自：http://www.cvchina.info/2010/03/02/depth-camera/

介绍两个深度摄像头，第一个是3DV systems的ZCam，第二个是PrimeSense公司的PrimeSensor。两个公司都是以色列的，他们的摄像头的性能也非常相近，都可以达到QVGA分辨率下60FPS，RGB image 和 Depth Image同步并逐点对应。但连个摄像头的结构大不相同，而且所使用的技术的名称也不一样，一个是TIME OF FLIGHT， 一个是LIGHT CODING.
1.ZCam
ZCam is a brand of time-of-flight camera products for video applications by Israeli developer3DV Systems. The ZCam supplements full-color video camera imaging with real-time range imaging information, allowing for the capture of video in 3D.

The original ZCam,[fn 1] released in 2000,[4] was an ENG video camera add-on used for digital video compositing.[5][6] Before agreeing in March 2009[7] to sell its assets to Microsoft,[8] 3DV had planned to release a ranging video webcam (previously called the Z-Sense), also under the name ZCam.[fn 1] The ZCam webcam was one of several competing real-time range imaging camera products in development that target home game controller applications.[fn 2]
———-摘自维基百科

2.PrimeSensor
PrimeSense is a fabless semiconductor company. Our technology empowers consumer electronic devices, such as TVs, set-top boxes, living-room PCs and more with natural interaction capabilities.
Our product, the PrimeSensor, contains the Reference Design and the NITE middleware.

The PrimeSensor Reference Design is a low-cost, plug and play, USB-powered device that can either sit on top of or next to a television screen or a monitor, or be integrated into them. The Reference Design generates realtime depth, color and audio data of the living room scene. It works in all room lighting conditions (whether in complete darkness or in a fully lit room). It does not require the user to wear or hold anything, does not require calibration and does not require computational resources from the host’s processor.
———摘自PrimeSense