This tool is designed specifically for PDF files that contain advertisement. The purpose is to clean up the bounding boxes drawn around all the objects contained in the PDF file, which can be either text or picture elements. There are three options that the user can utilize. The button 'Contained' eliminates all bounding boxes that are inside of other, obviously, greater bounding boxes. This occurs because in PDF files many images are put together by more than one image. For instance to achieve an effect of having a shadow, the image will have the item itself and then the shadow image. Other images can be choped into more than one object, to the visual appearence this makes no difference but inside the PDF file they are stored as 3 different images. The 'Merge Choped' option is designed to rid these.
The GUI contains two sliding bars at the top which determine the parameters for the action, buttons for different actions, and checkboxes which can blend out the boxes that are not needed to be shown. Each of the categories of boxes have a different color.
The button 'Merge Chopped' draws a yellow bounding box around boxes which are chopped off by other elements. The button 'Merge Overlapped' combines the bounding boxes that overlap each other by drawing a red bounding box around them. In the cases of 'Merge Chopped' and 'Merge Overlapped' the Deltas that is set by adjusting the sliding bars at the top of the panel are the values that are used as parameters for the merging. In the case of overlapping all boxes that are closer to each other than the Delta are merged. Hence, if the Delta is very low almost all boxes will be merged, if it is high very few boxes will be merged.
Example chopped:
Example contained:
Authors: Rob Kooper, Peter Bajcsy. Documentation: Peter Ferak.