Demo code for Imaging Class

Automated visual inspection systems are designed to perform quality control, inspection, and data collection. One or more imaging device captures photographs at various stages of the manufacturing process. These photos are analyzed to obtain important information, for example: rate of manufacture, defects in products, defects in packaging, robot guidance etc.   An example is the process of inspecting printed circuit boards. For a PCB to perform as designed, all components must be assembled correctly. This may include components that vary significantly in size: from fractions of a millimeter to a few centimeters. Inspecting these components is done with a range of methods from manual inspections to automated tests. There are a range of testing methods, including manual inspection, automated visual inspection, automated X-ray inspection, and in-circuit testing. A few articles on these methods are found here: [ 1 , 2 , 3 ]   In a typical manufacturing unit that generates say 40-5...

Cameras have become commodity products - with inexpensive cameras selling from $30 or so to expensive cameras that can cost thousands of dollars. No matter how good a camera, when a camera captures a real life object into an image, some distortion gets introduced. The slide deck here gives an overview of the type of distortions caused during image capture, and the theory behind  camera calibration. Here we will look at sample code for camera calibration. Calibration is the process of calculating the parameters that cause distortion. Remapping is the part where the parameters obtained from calibration can be used for a variety of purposes: Correct the distortion that is caused when a real life object is captured by a less-than-optimal camera. For example, fish eye lenses can distort an image. This can be corrected to some extent using camera calibration. Determine the real world coordinates (for example in millimeters or inches) corresponding to a certain object or measurement in t...