Scribd
Upload
172 views4 pages

Automatic Detection of Unexpected Accidents Under Bad CCTV

The document proposes a deep learning solution to automatically detect unexpected accidents in tunnels using CCTV footage. It would use an object detection and tracking system to 1) detect objects like cars, fires, and people in images and 2) track the movement of cars between frames to identify accidents. If an accident is detected, the traffic control system would be alerted to respond quickly and prevent further accidents.

Uploaded by

Roman Reings
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
172 views4 pages

Automatic Detection of Unexpected Accidents Under Bad CCTV

The document proposes a deep learning solution to automatically detect unexpected accidents in tunnels using CCTV footage. It would use an object detection and tracking system to 1) detect objects like cars, fires, and people in images and 2) track the movement of cars between frames to identify accidents. If an accident is detected, the traffic control system would be alerted to respond quickly and prevent further accidents.

Uploaded by

Roman Reings
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 4

Automatic detection of unexpected

accidents under bad CCTV


monitoring conditions in tunnels
The main causes of accidents in tunnels are

1. Wrong-Way Driving (WWD).


2. Stop.
3. Person out of vehicle in tunnel.
4. Fire.

Problems faced by it.


Due to unawareness of the accident there is an chance that maybe any other vehicle can
have a accident also the delay in any proper action can cost many lives.
Proposed Solution:
We propose a solution to this using a deep learning concept to detect the following accident
and also alert the traffic control system.
We will Use the concept or technique of ODTS (object detection and Tracking System) it will
accepts a video frame in time as an input to obtain Bounding Box (BBox) results by Object Detection
and compares the BBoxs of the current and previous video frames to assign a unique ID number to
each moving and detected object. This system makes it possible to track a moving object in time,
which is not usual to be achieved in conventional object detection frameworks.

Steps:-
1)The deep-learning-based video accident detection system uses a deep-learning object
detection network with the input of pre-processed still images as the first step. The deeplearning
model classifies objects such as a car, a fire, or person and regresses the bounding box, represented
by data in the form of a rectangular box.
2) The second step uses an object-tracking algorithm, simple online and real-time tracking
(SORT), which is based on the car object detected in the deep-learning object detection network.
The object-tracking process uses the car objects detected in the previous frame period and the
current frame period. When there is a pair of objects whose overlapping area ratio is equal to or
more than a predetermined value, the object number can be given and detected as the same object.
Flowchart:-

You might also like