We got the idea for this program when we thought of how a large amount of witness descriptions of criminals are rather inaccurate. In order to remedy this, we created mySentry, adding functionality such as automated police reporting to better the app so that it can help users who believe that they could be in danger or who witnessed a crime. After that, we added a second part of the app accessible only to the police force who can analyze the reports sent by the app, helping them create a more accurate map of crime in the area.
What it does
MySentry has two core user groups and it behaves differently depending on what kind of user is accessing the app. On the civilian side, the core function is the automated police reporting that takes a picture, analyzes it to see if there is a crime happening and if so, to describe the criminal, and send a report to the police with details about the crime and criminal. Users can also look up an area in the search bar and see user-reported crime statistics for the area. Lastly there is a button to directly call 911 if the user is in immediate danger.
For the police, the app aggregates the crime reports that are sent from the app, putting them into a readable format on a map. These crime reports come from both users and surveillance cameras that submit information to be analyzed.
How we built it
We used Clarifai for image processing and Twilio for sending alerts.
Challenges we ran into
It was difficult at first to work with the large data set and working with an unfamiliar technology such as Clarifai as we had not worked with image processing before. Also, when we realized that we could not give a video input into the program, we had to think creatively to figure out how to make it look like we are processing a video when we are really processing a series of images.