Re-purposing of content in order to utilize existing corporate assets in new ways is a great way to mine the value of those assets. One way to do this is to OCR text and put it up on search engines. However, drawings and images of products, parts, assemblies and maps must be carefully adapted to the web, so as to allow mouse actions with respect to particular objects on the images. These areas of sensitivity on the image are called hot spots. Users click on the hot spots to bring up more descriptions or to order items. Defining hot spots manually is error prone and very slow and costly because of the manual labor involved in the process. Furthermore, recognizing the part numbers or SKUs for entry into a database is also required.

Customers need a solution that can automatically:

Identify areas that are image objects
Identify areas that contain part numbers, SKUs or similar items

Where Inlite Research Fits In

ClearImage Tools offers the right set of capabilities to perform these tasks. It can:

Identify regions containing text or numbers of a given size
Identify regions containing images
Measure skew of regions of interest
Identify and remove lines where needed to improve the image
Remove margins and borders from the scanned page
Extract regions into new images, in order to show detailed views

With these capabilities it is possible to develop automatic extraction applications that perform the drudge work rapidly. Text regions can then be captured by OCR software. Recently, for example, a major automotive company created a maintenance application, to allow car technicians to look up assemblies, their sub assemblies and parts lists and to automatically order replacement parts and look up their costs. Their initial testing of the conversion process to hot spots showed about 10-30 minutes of work per page. They have over 100,000 pages. Clearly this was not going to pay off. With an imaging application built around ClearImage Tools, they were able to reduce the hot spot identification to less then a second per page, while the inspection process per page was down to 30 seconds. Now their project is underway because the entire project was completed in a few days.