Project Idea | (CAPTURED)

Last Updated : 11 Feb, 2022

Project Title : CAPTURED

Overview:
An android app which uses modern AI techniques to solve the one of the major security issue “THEFT” in India. Yes your own Android mobile would act as a guard to your house.

Specifications:
While thinking for the idea we found yourself digging around the problem of lack of cheaper and safer security solutions for our homes.
Hence we planned to use the latest technologies and expertise in the best way to make a project that would try to solve one of the biggest security problems of the society.
Then, we collaboratively came up with an idea of making an Android App which will detect unwanted person in the home with the help of Artificial Intelligence and will notify the owner in their absence.

How it works (Application) :
User have to first install the android app inside your android device and that’s the only major part 😛
After that just make a new account on the app and add the mobile number on which you want to get notified when a thief get CAPTURED .
Let’s call this number contactEmergency .
And that’s the only work you had to do. Now whenever you have to go somewhere and you are concerned about the Jewellery or cash , you just have to put your mobile device on a proper location so that it covers the entrance area of house, main room or whatever place at which you think the thief will definitely come. (Back or front camera of mobile should cover that location).
Now that’s it whenever a suspect will come to that location, your mobile will capture that and using IMAGE PROCESSING and ARTIFICIAL INTELLIGENCE it will notify you that a suspect has come to your house with the relevant picture of him. And now you can take immediate actions against him during the time he would be busy in executing the theft. 😛
He has been captured by CAPTURED. There is no way to escape.

Implementation:

Implementation to this app is a two step process:

At first our app will keep clicking the pictures using the mobile camera (placed at desired location ) recursively with a delay of some time period. Then it convert that image file into blob (binary form) and sends it to our server.
Now the server side work begins, which is the second and main step :
Now we have the blob image on our server. Now we are using image processing to process its different layers and identify whether there is a new human being comes in the picture as compared to the previous image (No Human) .
If No : No worry, It keeps doing the above two steps recursively.
If Yes : User will be notified using the push notification and also a message will be
sent to the mobile number which you have added in the emergency
contact (contactEmergency) saved in the app. And the message will include
the alert that “A Theft has been detected inside your house” and also the link
to the picture which is suspicious .

Tools Used:
1.Android studio for front-end of the android app.
2.Blob image conversion library to convert images into encrypted form for efficient exchange of images from
server to mobile device.
3.Image processing in layers using python library OpenCV and we have used special deep image analysis
algorithms those are mentioned in the algorithms section.
4.Server to device integration is done using PHP Laravel framework.
5. Dropbox API is used to store images with suspicious activities.
6. Firebase push notifications service is used to send push notifications to the owner
of the house.
7. msg91 api is used to send text sms in case of non-availability of internet connection
on owner’s mobile device.