Identification of Illegal Garbage Dumping with Video Analytics

We present a novel system for the identification of illegal garbage dumping from moving vehicles through the use of video analytics. The proposed system analyses a video sequence to detect garbage disposal from moving vehicles and identifies the owner of the vehicle based on the license plate of the vehicle. The system then moves to intimate the perpetrator of the act through a website. The offender may then pay the fine required through a payment portal available on the website. The proposed system comprises of modules for each of the stated tasks. The Vehicle Detection and Tracking module uses a Haar Cascade Classifier and Background Subtraction. The Garbage Identification module uses contouring and blob creation. The License Plate Reading module utilizes the OpenALPR library. The database interface accesses a Registration Certificate Book database and updates the website with relevant information. Due to the novelty of our problem statement, it was necessary to create our own dataset of video sequences featuring garbage of various sizes being thrown out of moving vehicles. The system successfully detects garbage in videos captured under versatile conditions, indicating that our approach is robust to changes in environment and illumination.