Video Recording and Analysis.

last updated on 11 May, 1997

While the modern video camera is very easy to use, most videos are made in a way that is not useful for sports analysis. This note is to give a few tips as to how to make the recording useful for quantitative analysis.

We're going to try to record the position as a function of time. This means that we must be able to register the position in an absolute way. To do this:

• The camera should be mounted on a stationary tripod and the sports action should take place with the camera beiling absolutely stationary. This is the easiest way to make measurements. You saw of this in the analyst of dropping balls, etc.
• For some actions, it's better to have a camera pan the action. The tripod should be carefully set up on a vertical axis so that the pan is in the horizontal plane. The vertical axis is necessary to make the vertical measurements accurate without a lot of fuss. For the horizontal calibration, there should be some visible markers on the g round. These can be used to calibrate the horizontal displacements.
• The camera should be set up a large distance [20 yards? ] from the action. Telephoto lens should be used. This will minimize the optical distortion of the photography process so that a clear position can be measured.
• The user should make verbal notes on the sound recording so that the later analyzer will know and understand the intent of the video shot.

Analyzing the video pictures. We now have a time series which is spaced by 1/30 of a second. To obtain a measurement of this, we must have the capability of single frame advance in the video playback. Unfortunately, the SONY cameras have a playback but no single frame advance. You must record the video onto a suitable single frame advance VCR. Professor Young has a Sharp video which does have the single frame cability.

To make the transfer, connect the video and sound cable from the camcorder to the vcr input. Be sure that the vcr is set to accept the video input.

The asymetrix AVP program can digitize video images. You must prepare a digital file for the recording first. Make a firm mental idea of what to digitize. It should be only about 1-2 seconds of video as the files generated are quite large and not sustainable in most computers. Record and stop.

The World in Motion system can read the .avi files prepared from the above process. You can further edit the video clip in length. The digitzation process MUST use all of the clip for the process. Be spare in your choice of frames. It's easy to get all clogged up with data so that the disk gets overfull.

Graphs can be generated from the world in motion program and displacement-time, velocity-time, and acceleration-time displayed. Be careful that the vertical axese are not automatically calibrated but must be massaged by hand. The default vertical axese can be used as arbitary axese and then hand calibrated after the reading.

up one level