Skip to content

Latest commit

 

History

History
7 lines (5 loc) · 620 Bytes

README.md

File metadata and controls

7 lines (5 loc) · 620 Bytes

DetectCarParkingSlot_Contest

QualifiedRound Contains qualified round proposed solution. The accuracy is hardly 10-20% using Yolo model prre-trained model.

FinalRound_ImprovedAccuracy_Functionality It detects car and empty slots in car park area with 90% accuracy. Used Yolo v3 model and PUCPR dataset to train Object detection model for Car parking area. Using trained model, collected all parking slots for Fully car parked image and similarly predict occupied car slot for interference. IOU (intersection over union) and NMS(non maximum suppression) technique used to find empty slots for inference image.