Using Object Detection on Social Media Images for Urban Bicycle Infrastructure Planning: A Case Study of Dresden

Abstract

With cities reinforcing greener ways of urban mobility, encouraging urban cycling helps to reduce the number of motorized vehicles on the streets. However, that also leads to a significant increase in the number of bicycles in urban areas, making the question of planning the cycling infrastructure an important topic. In this paper, we introduce a new method for analyzing the demand for bicycle parking facilities in urban areas based on object detection of social media images. We use a subset of the YFCC100m dataset, a collection of posts from the social media platform Flickr, and utilize a state-of-the-art object detection algorithm to detect and classify moving and parked bicycles in the city of Dresden, Germany. We were able to retrieve the vast majority of bicycles while generating few false positives and classify them as either moving or stationary. We then conducted a case study in which we compare areas with a high density of parked bicycles with the number of currently available parking spots in the same areas and identify potential locations where new bicycle parking facilities can be introduced. With the results of the case study, we show that our approach is a useful additional data source for urban bicycle infrastructure planning because it provides information that is otherwise hard to obtain.

Visualizing Point Density on Geometry Objects: Application in an Urban Area Using Social Media VGI

“…To make the visualizations reflect the results of the models presented in this paper in a meaningful way, we implemented a classification that considers the characteristics of the social media dataset used. In the dataset, more than half of images with bicycle detections contain only one bicycle, while merely 6.7 percent contain more than five (Knura et al 2021). Considering also that in the latter paper most of the 100 m × 100 m grid cells contained up to 35 parked bicycle detections on photos posted within the cells, we expected a similar pattern for bicycle detections in general.…”

“…VGI data As reference VGI data, we use the dataset from Knura et al (2021), where images containing bicycle detections in the area of Dresden, Germany, were retrieved from the Flickr YFCC100M dataset. The latter dataset was published under Creative Commons and contains 100 million user posts geolocated all over the world as photos or videos and related textual descriptions, covering the time span between 2004 and 2014.…”

“…While the representativeness, and therefore the fitness for use of a VGI data sample in a particular application has to be assessed based on the data itself (Senaratne et al 2017;Zhang and Zhu 2018), even representative VGI data could lead to overcrowded point clouds in very populated places when a large amount of points is georeferenced to a small area. For example, in the recent work of Knura et al (2021), the information on moving and parked bicycles in the city of Dresden was retrieved utilizing an object detection algorithm on social media images. While outside of the central city areas, one can easily identify each point feature that represents the location of the photo on which bicycles were detected, this is not possible in the central city areas, where point symbolization significantly overlaps (Fig.…”

Social Media Data in Urban Design and Landscape Research: A Comprehensive Literature Review

“…In urban design and landscape studies, Song et al [146] used YOLO v3 to perform the object detection of "Vehicle"," Bicycle", and" Pet" in High Line Park and the Atlanta Beltline. YOLO can also be used to predict multiple objects in the same image [147,148]. Moreover, facial recognition and expression detection are subparts of object detection, and the main object being detected is the human face from portrait social media images.…”

Vehicle Detection Using Machine Learning Model with the Gaussian Mixture Model (GMM)

“…The rapid growth and development of digital technology and the availability of video capture-based devices such as digital cameras and mobile phones with cameras are driving the explosive rise of network storage devices [1][2][3]. This is consistent with the nearly daily increase in the number of cars, but it is not accompanied by major volume changes [4].…”