-
Notifications
You must be signed in to change notification settings - Fork 2
Road Blocks
-
One of the initial roadblocks that we faced when dealing with the web scraping aspect of this project was figuring out how to login to someones Facebook page in order to begin scrapping our specified group for the data. We ended up finding the following source, Using Beautiful Soup, that helped us successfully login as Nicky to access the group.
-
Another roadblock that we encountered was Facebook updating their html. Our first solution for scrapping the group for the members' names and profile pictures worked for a few days, but when we went to fully implement it on the PYNQ board with the facial recognition, it failed to grab these items. Upon further inspection, the html from our Facebook group had changed and the tag classes had different names, so the source code had to be changed in order to read the new html.
-
Solved
- Automating the comparison part of the program. At first we had an image path and a name hard coded into the script to compare to the video feed. We automated this by using the Glob API which took a while to figure out how to use.
-
Unsolved
- Output lag from the PYNQ to the monitor. Processing and comparing the still image to the saved images takes time which creates lag when we output the video feed. We tried to fix this by processing less frames and by batching frames. Neither option made any difference in the lag.