With the increasing number of vehicles on campus, there arises a need to provide an automated system to provide students and staff with the information about the availability of parking space in a parking lot, thereby reducing the time taken to find empty parking spaces with minimal human intervention. In this project, we will be exploring options of solving the above stated problem using computer vision and machine learning techniques. We are using a video camera to record the vehicle movement in the parking lot. A set of videos are obtained from cameras set up at the entrances/exits of a parking lot. Using multiple approaches, we identify the number of cars entering/exiting the parking lot. These approaches include separating the data into training and testing data, feature extraction and machine learning techniques to predict the number of cars in a video. The results of the predictions are evaluated with the ground truth and the accuracy is obtained from the generated confusion matrix. To decrease the cost of production(implementation), an integrated vision-based system is an excellent choice.
risetarnished / vehicledetection Goto Github PK
View Code? Open in Web Editor NEWWith the increasing number of vehicles on campus, there arises a need to provide an automated system to provide students and staff with the information about the availability of parking space in a parking lot, thereby reducing the time taken to find empty parking spaces with minimal human intervention. In this project, we will be exploring options of solving the above stated problem using computer vision and machine learning techniques. We are using a video camera to record the vehicle movement in the parking lot. A set of videos are obtained from cameras set up at the entrances/exits of a parking lot. Using multiple approaches, we identify the number of cars entering/exiting the parking lot. These approaches include separating the data into training and testing data, feature extraction and machine learning techniques to predict the number of cars in a video. The results of the predictions are evaluated with the ground truth and the accuracy is obtained from the generated confusion matrix. To decrease the cost of production(implementation), an integrated vision-based system is an excellent choice.