Paper Reading #1: SpeckleSense: fast, precise, low-cost and compact motion sensing using laser speckle

Intro:
    Title: SpeckleSense: fast, precise, low-cost and compact motion sensing using laser speckle
    Reference Information: Research for this project took place primarily at the MIT Media Lab in Cambridge, MA, USA, but portions also took place at Comenius University DAI in Bratislava, Slovenia.
    Author Bios: 
Jan Zizka, Alex Olwal and Ramesh Raskar are a team of MIT researchers who have worked together on a few motion sense and also sound related projects together.

Summary:  

Technology such as smart phones and optical mouses have made highly sophisticated sensors and cameras very accessible and also very affordable. Laser speckle is basically the process of shining a coherent light source (or a regular laser passed through a diffuser) onto a surface or object and reading the resulting reflected light waves with an image sensor. The different light intensities result in a grainy, high-intensity image or "speckle".

Photos from SpeckleSense article

The research team tested the SpeckleSense motion sensing in a variety of different ways which are outlined and demonstrated below.

 

 

 

From SpeckleSense article

The team simply used the optical sensors in mice for the development of each of the prototypes they developed to demonstrate the usefulness of the SpeckleSense technology. The prototypes developed were TouchController, Mobile Viewport, and public display interaction.  The TouchController was created using an Apple Magic Mouse and adding the SpeckleSense technology.  It was implemented so that the multi-touch functionality could be combined with mid-air motions to manipulate a 3D viewer. The Mobile Viewport combined 3D tracking with mobile phones. An example of its use was in a photo viewer, where the phones distance from a surface controlled the zoom level of the image. Finally, SpeckleSense was used to improve public display interaction. Rather than a user being able to touch an interface or be required to mouse click on a screen, the system reads their hand motions and interprets them.

Related work: 
A method for processing laser speckle images to extract high-resolution motion - Andrew Houghton, Graham Rees and Peter Ivey (1997)
This paper discusses the use of laser speckle images to do motion detecting as well, but takes a pixel computation route rather than a light sensor. 

Robust technique of analyzing and locating laser speckle patterns for optical computer mice - Chih-Ming Liao, Ping S. Huang, Yi-Yuh Hwang, Ming Chen, Chung-Cheng (2009)
The difference here is obviously that this paper discusses the use of the laser speckle to improve the quality of optical mice. 

Minput: enabling interaction on small mobile devices with high-precision, low-cost, multipoint optical tracking - Chris Harrison, Scott E. Hudson (2010)
This is very much the same concept, but it is not quite as quick, as it implements optical sensors rather than the laser speckle. 

A laser speckle pattern technique for designing an optical computer mouse - P. Popov, S. Pulov, V. Pulov (2004)
The difference here is obviously that this paper discusses the use of the laser speckle to improve the quality of optical mice.

Statistical properties of laser speckle patterns - J.W. Goodman (1975)
This is simply a discussion of the laser speckle and not a motion sensing idea. 

Speckle in Ultrasound B-Mode Scans - Christoph B. Burkhardt (1978)
This paper discusses the laser speckle technology as used in Ultrasound image retrieval. 

Focus-position sensing using laser speckle - Joseph L. McLaughlin (1979)
This paper is simply discussing the ability to determine surface position with respect to focus.

Real time velocity measurement for a diffuse object using zero-crossings of a laser speckle - N. Takai, T. Iwai, T. Asakura (1980)
This paper discusses the use of the laser speckle to measure in-plane velocities rather than motion sensing capabilities. 

Accuracy in electronic speckle photography - M. Sjodahl (1997)
I am sure that the research from this paper was very useful in completing the motion sensing experiments of SpeckleSense. 

Visualization of retinal blood flow by laser speckle flowgraphy - H. Fujii (1994)
This is very similar to the Ultrasound implementation of the laser speckle. 

Evaluation:

The evaluation of SpeckleSense completed by the team was very minimal, and was qualitative and subjective. Five colleagues of the team were asked to complete the initial feedback evaluation. These colleagues were all males between the ages of 25 and 35 who had prior experience with motion controllers and mobile phones with multi-touch displays. These men were simply asked to interact with each of the prototypes and give and negative and positive feedback about the systems as they used them. The team plans to conduct more extensive user studies, but has yet to do so at this point.

Discussion:


I found this article to be absolutely fascinating. The fact that we are so very close to this 3D motion sense technology being in our pockets is amazing to me. I believe that the TouchController would be a phenomenal tool for giving presentations among other uses. The only thing that I found disappointing was the lack of variety in the test subjects for the evaluation of these prototypes. It would have been interesting to learn if the systems are intuitive and easy to use for the "average" user.