基于障碍物和车位检测的单阶段多任务YOLO-Parking算法研究

文章提出一种基于YOLOv4的端到端多任务网络模型用于自动泊车系统中的感知任务,以环视图像(around view monitor, AVM)作为网络输入,基于卷积网络提取图像特征信息,通过YOLO和DMPR-PS(directional marking-point regression-parking slot)检测头实现停车位与障碍物并行检测。在PS 2.0公开数据集上进行验证的结果表明,所提出的多任务检测方法能够同时检测停车位和障碍物,障碍物识别平均精度均值达到89.72%,车位识别查准率达到93.53%,网络检测速率为34.0帧/s,在满足自动泊车感知任务需求的同时提升了系统的检测效率。该文研究成果对自动泊车感知技术的发展具有一定的意义。

Strategies for multi-step-ahead available parking spaces forecasting based on wavelet transform

A new methodology for multi-step-ahead forecasting was proposed herein which combined the wavelet transform(WT), artificial neural network(ANN) and forecasting strategies based on the changing characteristics of available parking spaces(APS). First, several APS time series were decomposed and reconstituted by the wavelet transform. Then, using an artificial neural network, the following five strategies for multi-step-ahead time series forecasting were used to forecast the reconstructed time series: recursive strategy, direct strategy, multi-input multi-output(MIMO) strategy, DIRMO strategy(a combination of the direct and MIMO strategies), and newly proposed recursive multi-input multi-output(RECMO) strategy which is a combination of the recursive and MIMO strategies. Finally, integrating the predicted results with the reconstructed time series produced the final forecasted available parking spaces. Three findings appear to be consistently supported by the experimental results. First, applying the wavelet transform to multi-step ahead available parking spaces forecasting can effectively improve the forecasting accuracy. Second, the forecasting resulted from the DIRMO and RECMO strategies is more accurate than that of the other strategies. Finally, the RECMO strategy requires less model training time than the DIRMO strategy and consumes the least amount of training time among five forecasting strategies.

3D Bounding Box Proposal for on-Street Parking Space Status Sensing in Real World Conditions

Vision-based technologies have been extensively applied for on-street parking space sensing,aiming at providing timely and accurate information for drivers and improving daily travel convenience.However,it faces great challenges as a partial visualization regularly occurs owing to occlusion from static or dynamic objects or a limited perspective of camera.This paper presents an imagery-based framework to infer parking space status by generating 3D bounding box of the vehicle.A specially designed convolutional neural network based on ResNet and feature pyramid network is proposed to overcome challenges from partial visualization and occlusion.It predicts 3D box candidates on multi-scale feature maps with five different 3D anchors,which generated by clustering diverse scales of ground truth box according to different vehicle templates in the source data set.Subsequently,vehicle distribution map is constructed jointly from the coordinates of vehicle box and artificially segmented parking spaces,where the normative degree of parked vehicle is calculated by computing the intersection over union between vehicle’s box and parking space edge.In space status inference,to further eliminate mutual vehicle interference,three adjacent spaces are combined into one unit and then a multinomial logistic regression model is trained to refine the status of the unit.Experiments on KITTI benchmark and Shanghai road show that the proposed method outperforms most monocular approaches in 3D box regression and achieves satisfactory accuracy in space status inference.

Available parking space occupancy change characteristics and short-term forecasting model

Based on an available parking space occupancy （APSO） survey conducted in Nanjing, China, an APSO forecasting model is proposed. The APSO survey results indicate that the time series of APSO with different time-sections are periodical and self-similar, and the fluctuation of the APSO increases with the decrease in time-sections. Taking the short-time change behavior into account, an APSO forecasting model combined wavelet analysis and a weighted Markov chain is presented. In this model, an original APSO time series is first decomposed by wavelet analysis, and the results include low frequency signals representing the basic trends of APSO and several high frequency signals representing disturbances of the APSO. Then different Markov models are used to forecast the changes of low and high frequency signals, respectively. Finally, integrating the predicted results induces the final forecasted APSO. A case study verifies the applicability of the proposed model. The comparisons between measured and forecasted results show that the model is a competent model and its accuracy relies on real-time update of the APSO database.

Investigation of Ancient and Large Tree Resources in Green Spaces of Urban Parks: A Case Study of Yunqizhujing Park in Hangzhou, China

Conservation of ancient and large trees in domestic and overseas cities was compared, ancient and large trees were regarded as important cultural relics playing an important role in optimizing urban natural environment and enriching urban humanistic and natural landscapes, and they were also important contents of urban garden works symbolizing urban parks. A case study was carried out Yunqizhujing Park to study conservation of ancient and large trees in park green spaces of Hangzhou City, solutions to current problems were proposed, and constructive suggestions were given for the conservation of ancient and large trees in urban park green spaces.

Planning of Urban Public Disaster-prevention Green Spaces: A Case Study of Living Water Park in Chengdu City

Based on disaster-prevention functions of urban public green spaces, necessity and significance of emergency shelters in a city were analyzed. Taking Living Water Park in Chengdu for example, planning of urban public disaster-prevention green spaces was analyzed to further explore planning of urban disaster-prevention green space system and refugee spaces, and provide instructions for the planning and construction of urban disaster-prevention green spaces in China.

Evaluation of Recreation Service Quality of Slow Traffic Space in Urban Parks Based on IPA-Space Syntax Model

To build an“ecological and livable city”,it is important to figure out the relationship between human activity and urban park space and create a high-quality urban park through the integration of“people”and“park”.Taking Longhu Park in Huainan Province as an example,the study quantitatively analyzed and evaluated the characteristics and service quality of slow traffic space using importance-performance analysis(IPA)and space syntax,clarified the key indexes and areas to be optimized,and put forward corresponding design countermeasures and suggestions in combination with human activity demand.The results suggest that:①The mean performance of park recreation service quality is 3.63,less than the mean importance(3.95).Tourists pay special attention to the safety and convenience of slow traffic space,and the diversity and identiability of site functions is the focus“to be improved”.②The accessibility of Longhu Park shows a strong irregular shape in the southwest.The dead-end roads in the far lake area and the flora and fauna area on the northwest side have weak spatial perception,and the composite function carrying potential of slow traffic space is better reflected in a small range.Therefore,an optimization strategy for slow traffic space is proposed to enhance the continuity of road network,site identifiability,and functional diversification.The results will provide new thoughts of governance based on spatial humanism for the high-quality renewal of urban parks in the new era.

Study on the Distribution Characteristics and Influencing Factors of Park Green Space in Wuhan City

The distribution characteristics of park green space in Wuhan City were studied by standard deviation ellipse(SDE)method,average nearest neighbor index,kernel density estimation(KDE)method and scale index method.At the same time,the influencing factors of its distribution were discussed from three aspects of nature,social economy and culture with a geographical detector.The results show that(1)the distribution of park green space in Wuhan was relatively concentrated,and the directionality was not strong.The distribution mode was clustered mode.There were four main aggregation centers,which were located in Jianghan District,Qingshan District,Dongxihu District and Wuchang District respectively.Among them,the scale index of Jianghan District was the highest,up to 0.5902,while that of Xinzhou district was the lowest,only 0.0047.(2)The order of interpretation intensity of park green space distribution in Wuhan was population density(0.382)>regional GDP(0.278)>location of historical relics(0.165)>urbanization rate(0.136)>watershed zoning(0.069)>altitude(0.012).(3)The interaction of population density and watershed zoning had the strongest explanation for the distribution of park green space in Wuhan.Except for the influence of population density,the interaction of urbanization rate and regional GDP had the strongest explanation for the distribution of park green space in Wuhan.

Forecasting available parking space with largest Lyapunov exponents method

The techniques to forecast available parking space(APS) are indispensable components for parking guidance systems(PGS). According to the data collected in Newcastle upon Tyne, England, the changing characteristics of APS were studied. Thereafter, aiming to build up a multi-step APS forecasting model that provides richer information than a conventional one-step model, the largest Lyapunov exponents(largest LEs) method was introduced into PGS. By experimental tests conducted using the same dataset, its prediction performance was compared with traditional wavelet neural network(WNN) method in both one-step and multi-step processes. Based on the results, a new multi-step forecasting model called WNN-LE method was proposed, where WNN, which enjoys a more accurate performance along with a better learning ability in short-term forecasting, was applied in the early forecast steps while the Lyapunov exponent prediction method in the latter steps precisely reflect the chaotic feature in latter forecast period. The MSE of APS forecasting for one hour time period can be reduced from 83.1 to 27.1(in a parking building with 492 berths) by using largest LEs method instead of WNN and further reduced to 19.0 by conducted the new method.

Stereo garage parking space allocation model and simulation analysis

Based on grey neural network and particle swarm optimization algorithm,an automated stereo garage decision model is proposed to solve the problems of long waiting queue and low efficiency of automated parking garage.The gray neural network is used to forecast the stay time of the vehicle and particle swarm optimization algorithm is used to allocate the parking spaces in the stereo garage.The proposed stereo garage mathematical model is established on condition that vehicle arrival interval obeys Poisson distribution.The performance of stereo garage is evaluated by the average waiting time,average waiting queue length,average service time and average energy consumption of the customers.By comparing the efficiency indexes of the existing model based on near-distribution principle and the proposed model based on gray neural network and particle swarm algorithm,it is proved that the proposed model based on gray neural network and particle swarm algorithm is effective in improving the efficiency of garage operation and reducing the energy consumption of garage.

Design of Natural Children’s Recreational Spaces and Facilities in Urban Neighborhood Parks

Taking into account problems existing in urban children’s recreational spaces, this paper mainly discussed the method to design natural children’s recreational spaces and facilities, analyzed the characteristics and influencing factors of natural children’s recreational spaces and facilities based on the characteristics of behavioral and psychological cognition of preschool children, and concluded that children’s recreational spaces consist of five parts, namely, the border, entrance, internal activity space, viewing space and supporting space and that the design of recreational facilities should meet the four points of naturalness, appropriate fuzziness, appropriate scale and size, and diversities and opportunities, with a view to providing a reference for the design of children’s recreational spaces and facilities.

Recreational Behavior and Spatial Characteristics of Urban Parks Based on Multi-Source Information: A Case Study of Xiaohong Stone Carving Park in Nanjing

Urban parks are an important part of urban recreational spaces,and the development of analysis framework of recreational behavior and space is helpful for the design of urban parks to better meet the activity and usage requirements of visitors.In this paper,Xiaohong Stone Carving Park in Nanjing was taken as the research object to explore the relationship between landscape space and user behavior for design feedback.The orthographic image technology of unmanned aerial vehicle(UAV)and machine deep learning were used to analyze users behavior activities in the park,and the spatial characteristics of each node in the park were evaluated by semantic differential scale to explore the potential causes of behavior activities.The results showed that①there were significant differences in the number of visitors on holidays and working days,and young people and children were the main users.②The users activity of the park differed greatly on the inner and outer tour roads,and the users showed stronger willingness to stay on the inner tour roads.Recreational activities received more attention than cultural ones.③The overall distribution of users was significantly affected by hydrophilicity,and the recreational behavior and activity type were affected by multiple factors which include the hydrophilicity,safety,accessibility of cultural relics and the amount of recreational facilities.Finally,the enhanced strategies such as improving the cultural tour system,establishing hydrophilic safety belts and optimizing recreational facilities were put forward.

Exploration of Child-friendly Urban Park Design Strategies Based on Child Behavioral Psychology

By studying the behavioral psychology and needs of children,this paper analyzes the essence of child-friendly parks from the perspective of children,and discusses the design principles and key points of child-friendly parks,so as to provide a practical basis for the construction of child-friendly cities.

Fuzzy Logic Inference Applications in Road Traffic and Parking Space Management

In modern motoring, many factors are considered to realize driving convenience and achieving safety at a reasonable cost. A drive towards effective management of traffic and parking space allocation in urban centres using intelligent software applications is currently being developed and deployed as GPS enabled service to consumers in automobiles or smartphone applications for convenience, safety and economic benefits. Building a fuzzy logic inference for such applications may have numerous approaches such as algorithms in Pascal or C-languages and of course using an effective fuzzy logic toolbox. Referring to a case report based on IrisNet project analysis, in this paper Matlab fuzzy logic toolbox is used in developing an inference for managing traffic flow and parking allocation with generalized feature that is open for modification. Being that modifications can be done within any or all among the tool’s universe of discourse, increment in the number of membership functions and changing input and output variables etc, the work here is limited within changes at input and output variables and bases of universe of discourse. The process implications is shown as plotted by the toolbox in surface and rule views, implying that the inference is flexibly open for modifications to suit area of application within reasonable time frame no matter how complex. The travel time to the parking space being an output variable in the current inference is recommended to be substituted with distance to parking space as the former is believed to affect driving habits among motorist, whom may require the inference to as well cover other important locations such as nearest or cheapest gas station, hotels, hospitals etc.

About the Optimized Design of the Parking Space on the Campus of a College

In recent years, with the large increase in the number of motor vehicles in colleges and universities and the lag in campus planning, the relative shortage of parking spaces on campus has become increasingly serious. Taking Baoding College as an example, this article analyzes the current situation of static traffic on campus and finds out the problem of parking on campus through questionnaire surveys and field surveys. Analyze the growth trend of the number of motor vehicles based on the data, use the GM (1, 1) model and the linear fitting model to predict the number of motor vehicles in the future, and determine the size and layout of the parking lot based on the campus size, functional zoning, and road layout. The big campus-based parking system planning method based on big data can effectively solve the problems of small sample data, low accuracy, and poor timeliness of traditional methods, which improves the practicability and scientificity of planning results.

Post Occupancy Evaluation on Sociopetal Space of City Opening Park——Taking Jingxiu Park of Baoding as an example

Taking the sociopetal space with high using rate and welcomed by visitors in Jingxiu Park of Baoding as the study object,human demands on spatial characteristics and space factors were evaluated by the method of POE,and modification measures of the sociopetal space in city opening park were put forward,which provided evidence for solving regional problems of city park and creating urban open space.

Diversity Analysis on the Plant Communities of Urban Park Green Space in Wuhan

In order to systematically study the diversity of plant communities in park green space,78 plant communities of 10 representative parks in Wuhan City were investigated.The total 449 species belonging to 275 genera and 112 families were found in this investigation.The communities were divided into 35 dominant-species,13 life-forms and 6 vegetation-forms.The community structures and the composition of dominant species were also studied.The results showed that the shortages of the low diversity,high proportion of exotic species and simple community structures appeared in plant community of park green space in Wuhan City.It is proposed that species,especially native ones,should be abundant to keep a high level species diversity.At the same time,more simulated native communities should be utilized to build the natural,multi-layers,bright local vegetation characteristics landscape of urban parks.

Primary Study on the Index System of Underground Space Utilization in Urban Parks

The feasibility of underground space utilization in urban parks, the function and scale of underground space were explored, also the role, composition, quantification of the index system of underground space utilization in urban parks were analyzed, and the corresponding conceptual framework of index system were proposed.

Intelligent Parking Management System Based on Image Processing

This paper aims to present an intelligent system for parking space detection based on image processing technique. The proposed system captures and processes the rounded image drawn at parking lot and produces the information of the empty car parking spaces. In this work, a camera is used as a sensor to take photos to show the occupancy of car parks. The reason why a camera is used is because with an image it can detect the presence of many cars at once. Also, the camera can be easily moved to detect different car parking lots. By having this image, the particular car parks vacant can be known and then the processed information was used to guide a driver to an available car park rather than wasting time to find one. The proposed system has been developed in both software and hardware platform. An automatic parking system is used to make the whole process of parking cars more efficient and less complex for both drivers and administrators.

Impact of Ecological Corridor Construction on Urban Space-A Case Study of Fen River Park in Taiyuan City

With the city development and construction, ecological corridor has become one of the most attractive areas in urban districts. Ecological corridor has produced not only ecological benefits, but also economic benefits and social benefits, especially banded ecological park. By analyzing the impact of Fen River Park on leisure space, living space, communication space, commercial space of Taiyuan, and expounding shortcomings of constructing Fen River Ecological Corridor, countermeasures and suggestions were put forward.