Liu Jianhua

Prof. Liu Jianhua

Institute of Geomatics and Urban Spatial Informatics, Beijing University of Civil Engineering and Architecture, Beijing, China
orcid

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
10 January 2020 in Sensors
Verified
Journal Article
Open access
PDF file
SSShiran Song
Shiran Song
LJLiu Jianhua
Liu Jianhua
YLYuan Liu
Yuan Liu
GFGuoqiang Feng
Guoqiang Feng
HHHui Han
Hui Han
MDMingyi Du
Mingyi Du
High spatial resolution remote sensing image (HSRRSI) data provide rich texture, geometric structure, and spatial distribution information for surface water bodies. The rich detail information provides better representation of the internal components of each object category and better reflects the relationships between adjacent objects. In this context, recognition methods such as geographic object-based image analysis (GEOBIA) have improved significantly. However, these methods focus mainly on bottom-up classifications from visual features to semantic categories, but ignore top-down feedback which can optimize recognition results. In recent years, deep learning has been applied in the field of remote sensing measurements because of its powerful feature extraction ability. A special convolutional neural network (CNN) based region proposal generation and object detection integrated framework has greatly improved the performance of object detection for HSRRSI, which provides a new method for water body recognition based on remote sensing data. This study uses the excellent “self-learning ability” of deep learning to construct a modified structure of the Mask R-CNN method which integrates bottom-up and top-down processes for water recognition. Compared with traditional methods, our method is completely data-driven without prior knowledge, and it can be regarded as a novel technical procedure for water body recognition in practical engineering application. Experimental results indicate that the method produces accurate recognition results for multi-source and multi-temporal water bodies, and can effectively avoid confusion with shadows and other ground features.

2 January 2022 in International Journal of Remote Sensing
Verified
Journal Article
Closed access
YLYuan Liu
Yuan Liu
LJLiu Jianhua
Liu Jianhua
XNXiaohe Ning
Xiaohe Ning
JLJunjiao Li
Junjiao Li
The effective recognition and precise positioning of multiscale building rooftop is one of the key scientific problems that have yet to be resolved urgently in the current implementation of high-resolution remote sensing. In recent years, the automatic recognition of high-resolution image targets often employs convolutional neural networks to extract features. However, such traditional methods often ignore multiscale features of geographical objects, while lacking effective multiscale information extraction strategies. By utilizing the feature learning capability of deep neural networks, this study proposes a multiscale convolutional neural network named MS-CNN to recognize building rooftops from high-resolution remote sensing imagery. In addition, this study constructs a pedigree deep learning sample library based on the remote sensing Tupu theory that considers the spectral and geometric characteristics of building rooftops. Able to utilize feature segmentation mechanism and fusion enhancement strategy, MS-CNN enriches the receptive fields obtained by each convolution layer. The proposed network of this study is also compared with the famous Mask R-CNN method, proving the relative advantages of the MS-CNN method with multiscale characteristics. The experimental results show that the precision and recall metrics of the MS-CNN are 4.18% (.8655 vs. .8238) and 5.71% (.8380 vs. .7809) higher than those of the Mask R-CNN, respectively. The proposed method has been deployed in practical engineering projects in Vietnam and Myanmar, etc.

14 March 2025 in Sensors
Verified
Journal Article
Open access
PDF file
LJLiu Jianhua
Liu Jianhua
ZYZhijie Yang
Zhijie Yang
Sisi ZlatanovaSisi Zlatanova
Sisi Zlatanova
Songnian LiSongnian Li
Songnian Li
BYBing Yu
Bing Yu
Positioning information greatly enhances the convenience of people’s lives and the efficiency of societal operations. However, due to the impact of complex indoor environments, GNSS signals suffer from multipath effects, blockages, and attenuation, making it difficult to provide reliable positioning services indoors. Smartphone indoor positioning and navigation is a crucial technology for enabling indoor location services. Nevertheless, relying solely on a single positioning technique can hardly achieve accurate indoor localization. We reviewed several main methods for indoor positioning using smartphone sensors, including Wi-Fi, Bluetooth, cameras, microphones, inertial sensors, and others. Among these, wireless medium-based positioning methods are prone to interference from signals and obstacles in the indoor environment, while inertial sensors are limited by error accumulation. The fusion of multi-source sensors in complex indoor scenarios benefits from the complementary advantages of various sensors and has become a research hotspot in the field of pervasive indoor localization applications for smartphones. In this paper, we extensively review the current mainstream sensors and indoor positioning methods for smartphone multi-source sensor fusion. We summarize the recent research progress in this domain along with the characteristics of the relevant techniques and applicable scenarios. Finally, we collate and organize the key issues and technological outlooks of this field.

Verified
Journal Article
Open access
PDF file
LJLiu Jianhua
Liu Jianhua
JLJingyan Luo
Jingyan Luo
JHJiwei Hou
Jiwei Hou
DWDanqi Wen
Danqi Wen
GFGuoqiang Feng
Guoqiang Feng
XZXu Zhang
Xu Zhang
Accurate and fast indoor Location-Based Services (LBS) is very important for daily life and emergency response. Indoor map is the basis of indoor LBS. The model construction and data organization of indoor map are the key scientific problems that urgently need to be solved in the current indoor LBS application. In recent years, hybrid models have been used widely in the research of indoor map, because they can balance the limitations of single models. However, the current studies about hybrid model pay more attention to the model accuracy and modeling algorithm, while ignoring its relationship between positioning and navigation and its practicality in mobile indoor LBS applications. This paper addresses a new indoor map model, named Building Information Modeling based Positioning and Navigation (BIMPN), which is based on the entity model and the network model. The highlight of BIMPN is that it proposes a concept of Step Node (SN) to assist indoor positioning and navigation function. We developed the Mobile Indoor Positioning and Navigation System (MIPNS) to verify the practicability of BIMPN. Results indicate that the BIMPN can effectively organize the characteristics of indoor spaces and the building features, and assist indoor positioning and navigation. The BIMPN proposed in this paper can be used for the construction of indoor maps and it is suitable for mobile indoor positioning and navigation systems.

17 June 2019 in Remote Sensing
Verified
Journal Article
Open access
PDF file
SSShiran Song
Shiran Song
LJLiu Jianhua
Liu Jianhua
HPHeng Pu
Heng Pu
YLYuan Liu
Yuan Liu
JLJingyan Luo
Jingyan Luo
The efficient and accurate application of deep learning in the remote sensing field largely depends on the pre-processing technology of remote sensing images. Particularly, image fusion is the essential way to achieve the complementarity of the panchromatic band and multispectral bands in high spatial resolution remote sensing images. In this paper, we not only pay attention to the visual effect of fused images, but also focus on the subsequent application effectiveness of information extraction and feature recognition based on fused images. Based on the WorldView-3 images of Tongzhou District of Beijing, we apply the fusion results to conduct the experiments of object recognition of typical urban features based on deep learning. Furthermore, we perform a quantitative analysis for the existing pixel-based mainstream fusion methods of IHS (Intensity-Hue Saturation), PCS (Principal Component Substitution), GS (Gram Schmidt), ELS (Ehlers), HPF (High-Pass Filtering), and HCS (Hyper spherical Color Space) from the perspectives of spectrum, geometric features, and recognition accuracy. The results show that there are apparent differences in visual effect and quantitative index among different fusion methods, and the PCS fusion method has the most satisfying comprehensive effectiveness in the object recognition of land cover (features) based on deep learning.

4 July 2024 in IEEE Sensors Journal
Verified
Journal Article
Closed access
LJLiu Jianhua
Liu Jianhua
ZBZeng Baoshan
Zeng Baoshan
Songnian LiSongnian Li
Songnian Li
Sisi ZlatanovaSisi Zlatanova
Sisi Zlatanova
YZYang Zhijie
Yang Zhijie
WDWen Danqi
Wen Danqi
Scene-oriented multi-source sensor fusion for smartphone pervasive indoor localization is the key to location-based services (LBS), which is of practical significance to addressing the limitations of indoor navigation satellite signals and facilitating accurate location services within the final hundred meters. The rapid advancement of smartphone sensors and their performance provide a great opportunity for realizing smartphone indoor localization based on multi-source sensors. However, the limited adaptability of current localization methods hinders their widespread applicability, necessitating the development of a smartphone-based indoor localization method tailored for complex indoor scenes. This paper proposes a smartphone indoor localization method that integrates Map Location Anchors (MLAs) with Multi-sensor Fusion Location (MFL). The method identifies the sensor signal feature patterns of the smartphone’s built-in sensors in multi-scenes and binds them to MLA to serve map matching. The MLA is also utilized to correct the cumulative error of Pedestrian Dead Reckoning (PDR) to achieve indoor localization in multiple scenes by using fusion scheduling of different sensor modules. The experiment results show that the proposed method can achieve a localization accuracy of 1.01m in multi-floor scenes in collaboration with multi-source sensor localization modules matched with MLA, with high robustness and usability. The code of this paper is open source at https://github.com/GHLJH/MLA-MFL.