Dynamic detection of anomalies in pharmaceutical blisters using image processing

In today's field of medical science, we find people getting affected by multiple disorders, which are prone to long term diseases and cannot be cured without medication. This include solid medicines and liquid medicines i.e., tablets or syrups. In recent days we find tablet production by various vendors done is in ample, thus increasing the number of damages in the tablet which has to be verified at the stage of production itself. The damages caused can be of any form such as breakage, cracks or air blisters present in the tablet sheet which should be identified. Manual inspection of tablets sheet is a very challenging task. Thus we make use of median filtering and bounding box techniques in image processing for inspecting the defects in the tablets.