What And How: PeopleCounter-v1 API?
System Has Two Events:
- Computation (Done on a regular basis)
The system downloads the video, of the date we provide from the S3 bucket. Suppose we gave an i/p of 24-July-2019, the fetcher program fetches all the video names of that date from S3, puts them into a list, and loops it over the process of downloading and implementing the PeopleCounter Model. This looping operation also simultaneously adds the required records to the database. The schema of the first table looks like
vid_id | videoname | totalPeople | date | startTime
9 | 29-Feb-2019/ch02_20190724100000.mp4 | 7 | 2019–07–24 | 10:00:00
10 | 29-Feb-2019/ch02_20190724110159.mp4 | 7 | 2019–07–24 | 11:01:59
Once a particular date’s complete video is fetched, the system holds the potential to store the sum of PeopleCount of each video of that particular date.
id | date | totalPeople
7 | 29-Feb-2019 | 14
The “API”, is a medium to pull this data out from the DB and provide it further for other operations.
2. The API (For querying and transferring data)
So, PeopleCounter API, transfers the data stored in the database from the above operation, for further operations of analytics. For now, the API, takes the i/p as the date, to query the date in the database, and pull out all the respective data related to that date.
