GRADES 9-12
MOTION CAPTURE AND ANALYSIS

BACKGROUND INFORMATION: Motion Analysis As you look at your computer monitor the different tiny dots that makeup the picture on the screen are called pixels. An image has a "resolution" that is specified in number of pixels. For example, a resolution like 640 X 320 means there's 640 pixels in one direction and 320 in the next. The higher the resolution, the clearer the picture is. Video images from a digital camera have a set resolution that the manufacturer defines. Let's say that we have a picture of a piece of wood. If you take a ruler and measure the length of wood in the picture it measures 1.2 X .2 inches. I tell you that the image of the piece of wood has 120 X 20 pixels in it and also that the real piece of wood is 12 X 2 inches. Next to this piece of wood is a can of soda and although I know the measurement of the wood I do not know the measurements of the can of soda. But from the numbers above you can tell that the scale or ratio of the picture of the wood to the real piece of wood is 1:10. (Think of the scale on a road map - it might say 1 inch equals 50 miles.) If the can measured .3 X .8 inches in the picture then we know that the real can is 3 X 8 inches. That seems easy enough, but not sufficient to do motion analysis. Why? We weren't really concerned with lengths. The size of a tennis ball doesn't change in flight, nor does the length of Lindsay Davenport's foot or Venus William's racket or Todd Martin's head. We need to analyze motion - like velocity of the ball or the racket. Velocity = Distance/Time. A standard home video camera takes footage at 30 frames per second. That means every .033 seconds there was a frame shot (1 second/30 frames = .033 second/1 frame). So if we saw that the ball in the first frame was at one point and in the next frame it had moved 1 foot then we know that the velocity was 1 foot/.033 seconds = 30 feet/sec. But how do you mark these positions? Each position is in a different frame of the footage. Computer software can be used. This is how the software works. First we digitized the footage (the footage is read into the computer from a camera or tape deck and converted to a computer graphic). Here's where you had to be careful - the objects had to be in the same geometric plan. Let's go back to the piece of wood and soda can. The height of the can is 8 inches and the length of the wood is 12 inches. If they are in the same plane sitting next to one another all the scaling works correctly. But what it the can is three feet behind or three feet in front of the wood. In the picture when it is three feet behind the wood it would appear to be much small than 8 inches. You can compensate for all this but you have to be careful in your analysis and keep your frame of reference in mind.