Enhanced Facade Parsing for Street-Level Images Using Convolutional Neural Networks

Kong, Gefei; Fan, Hongchao

doi:10.1109/tgrs.2020.3035878

Cited by 21 publications

(18 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this paper, we used a pretrained model of YOLO v3 to extract windows, doors and balconies, which was trained on the Facade WHU dataset in Paris from Kong and Fan (2020). There are three subnetworks in Kong and Fan's work, and we only chose the window/door/balcony detection network.…”

Section: Extraction Results Of Facade Elementsmentioning

confidence: 99%

“…CNNs for semantic segmentation are not suitable for extracting them. Hence, we chose an object detection CNN, YOLO v3 (Redmon & Farhadi, 2018), for detecting facade elements (Kong & Fan, 2020). YOLO v3 is a one-stage, fast, and highly accurate object detection neural network.…”

Section: Detecting Facade Elementsmentioning

confidence: 99%

“…It can maintain a significant balance between detection accuracy and speed. As Kong and Fan (2020) have done, the Darknet53 was used as the backbone of YOLO v3. And the detection network was pretrained on our FacadeWHU dataset to detect windows, doors and balconies.…”

Section: Detecting Facade Elementsmentioning

confidence: 99%

See 2 more Smart Citations

An Interactive platform for low-cost 3D building modeling from VGI data using convolutional neural network

2021

Self Cite

View full text Add to dashboard Cite

The applications of 3D building models are limited as producing them requires massive labor and time costs as well as expensive devices. In this paper, we aim to propose a novel and web-based interactive platform, VGI3D, to overcome these challenges. The platform is designed to reconstruct 3D building models by using free images from internet users or volunteered geographic information (VGI) platform, even though not all these images are of high quality. Our interactive platform can effectively obtain each 3D building model from images in 30 seconds, with the help of user interaction module and convolutional neural network (CNN). The user interaction module provides the boundary of building facades for 3D building modeling. And this CNN can detect facade elements even though multiple architectural styles and complex scenes are within the images. Moreover, user interaction module is designed as simple as possible to make it easier to use for both of expert and non-expert users. Meanwhile, we conducted a usability testing and collected feedback from participants to better optimize platform and user experience. In general, the usage of VGI data reduces labor and device costs, and CNN simplifies the process of elements extraction in 3D building modeling. Hence, our proposed platform offers a promising solution to the 3D modeling community.

show abstract

Section: Extraction Results Of Facade Elementsmentioning

confidence: 99%

Section: Detecting Facade Elementsmentioning

confidence: 99%

See 1 more Smart Citation

An Interactive platform for low-cost 3D building modeling from VGI data using convolutional neural network

2021

Self Cite

View full text Add to dashboard Cite

show abstract

Section: Methodsmentioning

confidence: 99%

“…The presented work is based on our two earlier works (Kong and Fan 2020;Fan et al 2021). The first one (Kong and Fan 2020) proposed a deep learning method to detect façade elements from low-quality street-level images. The second one (Fan et al 2021) introduced the first version of VGI3D from the perspective of geographic information system (GIS).…”

Section: Methodsmentioning

confidence: 99%

VGI3D: an Interactive and Low-Cost Solution for 3D Building Modelling from Street-Level VGI Images

2021

Self Cite

View full text Add to dashboard Cite

Applications in smart cities are inseparable from the usage of three-dimensional (3D) building models. However, the cost of generating and constructing 3D building models with semantic information is high both in time and in labour. To solve this problem, we developed a web-based interactive system, VGI3D, with the ambition of becoming a VGI platform to collect 3D building models with semantic information by using the power of crowdsourcing. VGI3D is a platform-independent software program that is composed of a spatially relational database (PostgreSQL/PostGIS) for the storage and management of spatially geometrical data and other software modules, allowing users to import, analyse, reconstruct, visualise, modify and export 3D building models according to the OBJ/CityGML standard. In this paper, we present the VGI3D in detail, focusing on relevant technical implementations, and report the results of limited usability testing aimed at optimising the system and user experience. After limited expert and non-expert participants’ testing, we proved the usefulness of VGI3D and its promising value for the 3D modelling community.

show abstract

Chaotic Image Encryption: State-of-the-Art, Ecosystem, and the Promise of Quantum-Inspired AI

Zolfaghari,

Nemati,

Yanai

et al. 2023

Crypto and AI

View full text Add to dashboard Cite

Enhanced Facade Parsing for Street-Level Images Using Convolutional Neural Networks

Cited by 21 publications

References 41 publications

An Interactive platform for low-cost 3D building modeling from VGI data using convolutional neural network

An Interactive platform for low-cost 3D building modeling from VGI data using convolutional neural network

VGI3D: an Interactive and Low-Cost Solution for 3D Building Modelling from Street-Level VGI Images

Chaotic Image Encryption: State-of-the-Art, Ecosystem, and the Promise of Quantum-Inspired AI

Contact Info

Product

Resources

About