Fully automated roadside parking spot detection in real time with deep learning. (14th September 2020)

• Record Type: Journal Article
• Authors: Çiçek, Emre; Gören, Sezer
• Other Names: Awan Irfan guestEditor.; Younas Muhammad guestEditor.; Benbernou Salima guestEditor.; Ogiela Lidia guestEditor.; Leu Fang‐Yie guestEditor.; Fiore Ugo guestEditor.
• Is Part Of: Concurrency and computation. Volume 33:Number 23(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-09-14
• Subjects: convolutional neural network -- faster region‐based convolutional networks -- fully convolutional networks -- object detection -- parking spot detection -- road segmentation; Parallel processing (Electronic computers) -- Periodicals; Parallel computers -- Periodicals; 004.35
• DOI: 10.1002/cpe.6006
• Languages: English
• ISSNs: 1532-0626

Abstract

Summary: Searching for a roadside parking spot in crowded cities is a burden. In this article, we propose a vision‐based mobile cloud parking management solution, which is fully automated such that it can find roadside parking spots in any street with flowing traffic. To develop such a system, we employ both object detection and road segmentation methods. Thus, we do not need to manually label and train for every distinct street and do not mark out parking spot boundaries and surrounding road areas. In our approach, fully convolutional networks, specifically the FCN‐VGG16 model and KITTI road dataset are used for road segmentation, whereas Faster Region‐based Convolutional Neural Networks and Microsoft Common Objects in Context dataset for object detection. Our Road Boundaries algorithm automatically identifies the road polygon excluding the roadside. The main contribution here is to differentiate the parked cars from the moving cars on the road and detect the available parking spot between the parked cars on the roadside and direct the driver to the nearest spot. On GPU, we achieved a frame rate of 1.5fps and up to 83% accuracy with flowing traffic and 92% with no traffic flow. These results promise a potential solution on a city‐wide scale.