In order to comprehend what Image Processing is without going into individual algorithms, consider this simple application example:
The tiny components seen in the image above come to a machine in small compartments of a component tape as seen in image below. It is essential for the machine that will use the components to know that they are all placed the same way. A more advanced machine could handle that the components are NOT placed the same way as long as it knows HOW each component is oriented.
Only components B, D, and H are placed correctly. Compartments A and F are empty and component C is placed upside-down. Components E and G are located within their compartments, but facing left and right respectively.
A digitalized 3D-image of the component tape could look like above. Application specific software could use the height data from the image to determine where the holes and compartments are located. That information could be used to calculate areas around the top, center, and bottom of each compartment. Then, pixels in certain height intervalls could be counted or colored within the different areas. Finally, each compartment can be classified and the result communicated to the machine.
For instance, in the processed image above too many pixels colored red means that the component is either missing or facing to the side. A component that doesn't have enough aqua colored pixels at the top and bottom of the compartment is either placed upside-down or facing to the side. In order to know that a component is facing left both these tests are done and then the center of the red pixels has to be to the right of the center of the compartment.
Please, consider us for your next Factory Automation project.
sales@3D-Integration.com