Parking-stall vacancy indicator system, based on deep convolutional neural networks

Abstract

Abstract-Parking management systems, and vacancyindication services in particular, can play a valuable role in reducing traffic and energy waste in large cities. Visual detection methods represent a cost-effective option, since they can take advantage of hardware usually already available in many parking lots, namely cameras. However, visual detection methods can be fragile and not easily generalizable. In this paper, we present a robust detection algorithm based on deep convolutional neural networks. We implemented and tested our algorithm on a large baseline dataset, and also on a set of image feeds from actual cameras already installed in parking lots. We have developed a fully functional system, from server-side image analysis to frontend user interface, to demonstrate the practicality of our method.

A Survey on an Emerging Area: Deep Learning for Smart City Data

“…The work in [17], [68], [69] applied multi-task learning to jointly learn people's movement and transportation mode patterns. Often when a training dataset may be too specific or small to learn a good model from scratch, transfer learning can be applied by pre-training a model with large available dataset and then fine-tuning the model with the data for specific tasks such as fire detection [70], parking lot detection [71], and crisis-related tweets classification [56].…”

“…The system was deployed at the parking lot of the research campus in Pisa as part of the smart city application. Similarly, the work in [71] presented a robust parking lot detection algorithm based on a light version of VGGNet. It was used to report the occupancy status of a parking stall based on the images taken by cameras.…”

“…This characteristic makes transfer learning particularly attractive for more efficient data analytics. In fact, it has been applied in a limited number of small-scale studies such as fire detection [70] and parking lot detection [71]. In the future, we would expect extensive applications of transfer learning in many other smart city domains.…”

Um Mapeamento Sistemático sobre Sistemas que envolvem IoT e Deep Learning conjuntamente

“…O Quadro 4 mostra o mapeamento dos sistemas e aplicações já existentes retornados pela pesquisa. Sistema para robôs inteligentes 1 [2] Sistemas de orientação e informação de estacionamento 1 [11] Sistema de reconhecimento, em tempo real, de objetos em uma imagem 1 [12] Sistema de reconhecimento de atividades humanas 1 [13] Sistema de raciocínio e inferência de situações 1 [15]…”

IoT -Based Approach to Monitor Parking Space in Cities

“…Using machine learning has been the next thing in today's world, which implies to unusual pattern in recognizing the parking spaces being allocated and dynamically checking out the vacancy in the parking lot area [9]. The sensors that can be used to achieve the task and get back with the information on the parking lot area are, ultrasonic, inductive loops, infrared sensors as the sensors tend to be very reliable but for huge parking lots the cost can be huge and hence using convolutional neural network can be an option to get the desired output [13]. The process and other related measures take enough power and capacity, but this can be resolute by connecting the network in the hub which in here acts in the cloud [14].…”