Algorithm

Divide to three, Python scripts

1. FrameCapture
    a. Open stream
    b. Catch 1 frame every n=1 seconds
    c. Save frame in folder
        i. /data/raw_new/raw_<camID>_<timestamp>.jpg
2. FrameAnalysis
    a. Open first photo from /data/raw_new/
    b. Load AboutCam json
    c. OpenCV:
        i. Count people
        ii. Draw rectangles on peoples
        iii. Prepare json EventCam
    d. Save:
        i. Save EventCam to /data/analyzedJson_new/analyzed_<camID>_<timestamp>.json
        ii. Save new picture with rectangles to /data/analyzedPhoto_new/analyzed_<camID>_<timestamp>.jpg
        iii. Move raw picture from /data/raw_new/raw_<camID>_<timestamp>.jpg to /data/raw_archive/raw_<camID>_<timestamp>.jpg
3. FrameSend
    a. Open MongoDB connection
    b. Select database and collection
    c. Send all files from folders, and move them:
        i. /analyzedPhoto_new/ and next move to /analyzedPhoto_archive/
        ii. /analyzedJson_new/ and next move to /analyzedJson_archive/