Pixel mapping is controlling an array of lighting fixtures by playing images or video. It is typically used to sendDMX to LED panels but can be used to control any number of fixtures in any configuration. Each addressable fixture is assigned to a position in a grid. Playing back an image or video translates the color and intensity of the pixel into DMX for the fixture.
At the most basic level think of the points of light as an individual pixel. I've seen it done with individual source 4s, the denser the "pixels" the more detail you can achieve, think of LED screens and digital signs at ballparks or on hotels, same concept, but using stage lighting.
In the video world pixel mapping relates to changing between or matching digital resolutions, for example mapping a 1366x768 image to a 1920x1080 native display. In that case both resolutions are 16:9 format but 1:1 pixel mapping would only use a portion of the display and filling the display with the image requires a much more complex process to 'remap' the original 1366 pixel by 768 pixel resolution image to a 1920 pixel x 1080 pixel resolution image.
I believe that pixel mapping may also have a slightly different application in the digital photography world.