Monthly Archives: September 2009

compression library

For my multi-camera project, I need to use some compression library to round the 2G user application space limit, otherwise I can capture the data for only 1~2 secs before all the memory is filled up. I have tried to switch to 64 bit system before, but the camera driver didn’t work stably. Today I wrote a test program using a very good lossless compression library – Zlib. I thought this would meet my expectation, but the test compression ratio is only 1.7:1 for a typical scene. I guess I have to try some lossy compression library, like jpeg image compression library, this would definitely suit my case.

Independent JPEG Group
Where can I get image compression program?

3D body scanner

I have seen several types of 3D scanners for our face recognition projects before. Most of them are based on laser technologies. The subjects usually need to sit still for a couple of seconds for the device to collect enough data samples. Here comes a new one – Intellifit Virtual Fitting Room™ (VFR) – for consumer markets.

The following materials are copied from their website.


The VFR sends safe, low power radio waves toward the customer’s fully clothed body and, in about ten seconds, collects numerous data points that are used to calculate their accurate body measurements. The “vertical wand” in the Intellifit Virtual Fitting Room™(VFR) contains 196 small antennas that send and receive low power radio waves. In the fifteen seconds it takes for the wand to rotate around a fully clothed individual, the radio waves send and receive low power signals.

The signals don’t “see” a customer’s clothing but reflect off their skin which is, basically, water. The signals are similar to cell phone signals but less than 1/350th of the power of those signals and they do not penetrate the skin.

When the wand’s rotation is complete, Intellifit has recorded over 200,000 points in space, basically x, y, and z coordinates of where that customer is standing. Intellifit software then electronically measures the “point-cloud”, producing a file of dozens of accurate body measurements; the raw data is then discarded.