Vehicle Identification Model

Computer Vision

Make the daily usage data of the various car wash slots accessible and understandable. The aim is to record customer trends and monitor the correlation between footfall, positioning, vehicle type and weather.

Description of activities

Development of an ETL pipeline in Azure environment for data collection via REST API from cameras in the service, processing and saving videos in cloud environment. Videos are processed by the Computer Vision model every day to populate a partitioned table on a daily basis with information about vehicles recognized directly from the video. Daily scheduling of the automatic data update job.

Tools and technologies used

Azure, Yolo, Python, Streamlit, Bitbucket