We are prototyping some setups to be able to capture data of 7-segment displays (think LCD watch) that come in major medical devices such as glucometers and blood pressure cuffs. One way of obtaining the data is by cracking open the unit and sniffing out the electronic feeds. Another way is by purchasing units that already do this. Most of those types of machines don’t share the data openly though. So last night, our lab neighbor, Shane Colton suggested that we could use a webcam and image recognition software.
I like this idea and we’ll be implementing in the next few days. We’re looking for a few brave OpenCV experts to help.
Why do this? Because it takes legacy systems