Overview / Usage

vCheck aims to bring about changes in how security is handled across various segments such as Special Economic Zones (SEZ), Hotels, etc. Using object recognition & heat mapping via Deep Learning (Convolutional neural networks & Tensorflow), we intend to automate & secure the entire security process.

Methodology / Approach

Camera detects the number plate & classifies it using an OCR classifier (Neural Network) for recording the number plate of the vehicle approaching it.
Once the vehicle stops for ID verification and/or verification of the interior of the trunk, the camera beneath the car will flag dangerous or uncharacteristic objects using an object detection model and our own classifier trained on custom data set in tandem.
On the fly, all this data would be stored in a Database.
A similar setup will be installed at the exit gate to record the vehicle exiting, again using the OCR classifier & adding it in the existing database entry.

Technologies Used

Intel Optimized Libraries
OpenVINO
Tensorflow
IP/RealSense Cameras

Collaborators