|
Image Alignment
When comparing the stored image to the live image for purposes of alignment, highlighting the differences in alignment is the goal of the image overlay. When comparing the stored image of the golden standard to the live image of the part under test for verifying integrity the two images must be in alignment. This can be done mechanically, but it is also possible to shift the image in software. One method could be implemented by performing a correlation operation between the two images. It is possible to compare one image relative to the other while shifting the data of one image in X and Y, using software, until the peak correlation is detected. The source image can then be shifted using the SRCCOPY function with shifted X and Y origins. By viewing the resulting image overlay of a comparison of two images that have been aligned to each other, the remaining differences reflect the production errors only. Once the images have been aligned to each other the image overlay can be adjusted through threshold and inverting to bring out the important differences from missing components, solder bridges etc.
Conclusion
There are many inspection tasks requiring the comparison between a golden standard and a production part. By using an image capture system to acquire an image of the standard and store it, numerous manipulations can be made to highlight differences between the live image of a production part and the stored image of the standard. The Microsoft Raster Operations are one way to manipulate those images to display the discrepancies of the part under inspection. By creating pseudocolor overlays on monochrome images that are generated from differencing algorithms, errors can be highlighted in an obvious manner.
Page 1 2 3 4 5 6 |
|
