Here’s an fascinating experiment that attempts to measure the top quality of a linear rail by using a form of visible odometry, achieved by mounting a digicam on the rail and examining the video with open-supply software program usually applied to stabilize shaky movie footage. No linear rail is best, and it ought to be probable to evaluate the degree of imperfection by recording movie footage while the digital camera moves down the length of the rail, and examining the outcome. Imperfections in the rail should cause the video clip to sway a proportional amount, which would let just one to characterize the rail’s quality.
To take a look at this thought, [Saulius] hooked up a higher-definition digital camera to a linear rail, pointed the digicam towards a large-distinction textured pattern (generating the resulting movie a lot easier to assess), and recorded online video even though shifting the digital camera throughout the rail at a set velocity. The resulting movie receives fed into the Deshaker plugin for VirtualDub, of which the essential section is the
deshaker.log file, which includes X, Y, rotate, and zoom correction values essential to stabilize the movie. [Saulius] made use of these values to produce a graph characterizing the linear rail’s top quality.
It is a intelligent evidence of notion, in particular in how it makes use of no distinctive applications and leverages a video clip stabilizing algorithm in an abnormal way. Nonetheless, the results are not exactly simple to transform into concrete, authentic-planet measurements. Turning image outcomes into micrometers is a matter of counting pixels, and for this undertaking video stabilizing is an imperfect device, considering the fact that the algorithm prioritizes visual results alternatively of absolute measurements. Continue to, it is an attention-grabbing experiment, and perfectly capable of measuring rail high quality in a relative feeling. Cannot assist but be a bit curious about how it would profile some thing like these cardboard CNC modules.