Jisoo Park | Indiana State University (original) (raw)

Uploads

Conference Presentations by Jisoo Park

Research paper thumbnail of Resilient Fire Prevention and Management Strategies for Structures and Materials Stored under Urban Bridges

The collapse of an urban bridge resulting from a massive fire causes significant economic damage.... more The collapse of an urban bridge resulting from a massive fire causes significant economic damage. Preventing a recurrence of such accidents involving massive fire requires special management regulations and strategies for the fires that break out under bridges. This study introduces several strategies for the fire risk management of materials stored under bridges through the lessons learned from similar cases in South Korea. In addition, this study investigates the optimum fire management method for the space under bridges based on the best practices conducted in Korea. This study also conducted fire simulations for materials temporarily stored under bridges with different heights and volumes of the materials. Based on the results of the fire simulation, this study establishes a required distance between bridges and facilities or materials under bridges. Finally, this study suggests a guideline to establish fire management standards for the materials and facilities under bridges.

Papers by Jisoo Park

Research paper thumbnail of Deep learning-based UAV image segmentation and inpainting for generating vehicle-free orthomosaic

International Journal of Applied Earth Observation and Geoinformation, 2022

A low-altitude orthomosaic derived by an unmanned aerial vehicle (UAV) has been widely utilized f... more A low-altitude orthomosaic derived by an unmanned aerial vehicle (UAV) has been widely utilized for various purposes in large-scale infrastructure management. However, unwanted objects, such as cars and trucks, captured in the aerial images captured by the UAV have negative impacts on the quality of orthomosaic. To this end, this study presented a novel method to remove the effect of unwanted objects on UAV-generated orthomosaic. The proposed method applied a deep learning-based image segmentation and inpainting algorithm to remove the vehicles from individual UAV images before processing structure from motion (SfM), and then it resulted in generateing an orthomosaic with the inpainted UAV images. To validate the proposed method, this study conducted a case study in actual highway environment and compared the performance of the proposed method with that of another method, which directly removes and inpaints vehicles from the final orthomosaic. Through comparison tests, it is shown that the proposed method is more effective than the other. The proposed automatic vehicle-free orthomosaic generation method can contribute to creating up-to-date immersive content for transportation infrastructure management.

Research paper thumbnail of Assessment of Construction Points for Grade Control and Reference in 3D

The objective of this research is to investigate the causes of inaccuracies of grade elevations f... more The objective of this research is to investigate the causes of inaccuracies of grade elevations from current practices and identify best practices for grade control and referencing on transportation infrastructure construction projects in Georgia. A nationwide questionnaire survey was conducted with contractors, engineers, surveyors, and owners, and the following causes of the discrepancies are identified: (1) long time gap between design and construction phases; (2) poor techniques in surveying; and (3) long interval between survey points. Through an actual case study with an ongoing highway construction project, the following conclusions and recommendations are made: (1) a mandatory field survey before earth moving is recommended; (2) unmanned aerial vehicle (UAV) and Mobile Mapping System (MMS) are effective for 3D grade control in planning and site cleaning phases; (3) terrestrial laser scanning (TLS) is appropriate for 3D grade control in a subgrade grading stage; and (4) robotic total station (RTS) is recommended for the final grading stage.

Research paper thumbnail of CNN-Based Person Detection Using Infrared Images for Night-Time Intrusion Warning Systems

Sensors

Night-time surveillance is important for safety and security purposes. For this reason, several s... more Night-time surveillance is important for safety and security purposes. For this reason, several studies have attempted to automatically detect people intruding into restricted areas by using infrared cameras. However, detecting people from infrared CCTV (closed-circuit television) is challenging because they are usually installed in overhead locations and people only occupy small regions in the resulting image. Therefore, this study proposes an accurate and efficient method for detecting people in infrared CCTV images during the night-time. For this purpose, three different infrared image datasets were constructed; two obtained from an infrared CCTV installed on a public beach and another obtained from a forward looking infrared (FLIR) camera installed on a pedestrian bridge. Moreover, a convolution neural network (CNN)-based pixel-wise classifier for fine-grained person detection was implemented. The detection performance of the proposed method was compared against five conventiona...

Research paper thumbnail of Point Cloud Information Modeling: Deep Learning-based Automated Information Modeling Framework for Point Cloud Data

Journal of Construction Engineering and Management, 2021

Since point clouds have reality measurements of physical objects, they are often used to reconstr... more Since point clouds have reality measurements of physical objects, they are often used to reconstruct the as-built 3D model of building construction sites through a modeling process called Scan-to-BIM. However, the reality measurements in point cloud data, such as actual color and deformed shapes of the original objects, could disappear during the solid modeling process. In addition, the conventional Scan-to-BIM pipeline still requires significant time and manual effort for object classification and shape representation. To address these problems, this study proposes a novel information modeling framework for point clouds, called point cloud information modeling (PCIM). PCIM can automatically recognize construction objects and their properties with deep learning approaches. Furthermore, it can store information in the original point cloud data with a hierarchical structure, rather than converting it to a solid or rigid model. To validate the overall PCIM concept, this research conducted a case study with an actual building construction project. The test results demonstrate that PCIM can be an effective tool for the as-is information modeling of structures and facilities during construction.

Research paper thumbnail of Framework for Automated Registration of UAV and UGV Point Clouds Using Local Features in Images

Automation in Construction, 2019

Automatically registering 3D point clouds generated by unmanned aerial and ground vehicles (UAVs ... more Automatically registering 3D point clouds generated by unmanned aerial and ground vehicles (UAVs and UGVs) is challenging, as data is acquired at different locations with different sensors, consequently resulting in different spatial scales and occlusions. To address these problems, this study proposes a framework for the automated registration of UAV and UGV point clouds using 2 2D local feature points in the images taken from UAVs and UGVs. This study first conducted field experiments by varying the angles of the UAV camera to identify the optimal angle with which to detect sufficient points matching with the images taken by the UGV. As a result, this study identified that a combination of UAV images taken at 30˚ and 90˚ is appropriate for generating a sufficient number of matching points and attaining a reasonable level of precision. The UAV and UGV point clouds were initially scaled and registered with a transformation matrix computed from the 3D points corresponding to the 2D feature matching points. The initially aligned point clouds were subsequently adjusted by the Iterative Closest Point (ICP) algorithm, resulting in the root mean square error (RMSE) of 0.112 meters. This promising result indicates that full automation of spatial data collection and registration from a scattered environment (e.g., construction or disaster sites) by UAVs and UGVs is feasible without human intervention.

Research paper thumbnail of As-is Geometric Data Collection and 3D Visualization through the Collaboration between UAV and UGV

the 36th International Symposium on Automation and Robotics in Construction (ISARC), 2019

The characteristics of dynamic construction sites increase the difficulty of collecting the high-... more The characteristics of dynamic construction sites increase the difficulty of collecting the high-quality geometric data necessary to achieve construction management activities. This paper introduces a new autonomous framework for 3D geometric data collection in a dynamic cluttered environment using an unmanned ground vehicle (UGV) and an unmanned aerial vehicle (UAV). This method first deploys UAV to collect photo images of a site and builds a point cloud of the 3D terrain of the site, including obstacle information. A mesh grid is then created from the UAV-generated point cloud, and simulation for laser-scan planning is conducted to determine the stationary laser-scan positions at which a mobile robot can collect data with less occluded views while capturing all crucial geometric information. Finally, optimal paths for the UGV to navigate among the estimated scan positions are generated. Promising test results regarding data accuracy and collection time show that the proposed collaborative UAV-UGV approach can significantly reduce human intervention and provide technologies for enabling construction site to be frequently monitored, updated, and analyzed for timely decision-making.

Research paper thumbnail of UAV-assisted autonomous mobile robot navigation for as-is 3D data collection and registration in cluttered environments

Automation in Construction, 2019

This paper discusses enabling autonomous mobile robots to operate in unstructured terrain and env... more This paper discusses enabling autonomous mobile robots to operate in unstructured terrain and environments and gather data in a time-efficient manner. In rugged environments such as construction sites and disaster sites, spatiotemporal data is difficult to acquire since these environments are potentially hazardous to humans. Autonomous robot systems represent a reasonable solution to provide topographic and surveillance data that can assist human activity in these environments, whether for exploration, mapping, or search and rescue. However, automatically operating the mobile laser scanning robots at cluttered environments is challenging because as-is geometrical conditions of the site are difficult to comprehend from the ground level due to the blocked line-of-site views. To address the issue, this paper introduces a new framework for operating mobile robots equipped with a laser scanning system in cluttered outdoor environments with the aid of an unmanned aerial vehicle (UAV). To obtain an initial map from the current field, this method first deploys UAV to collect photographic images of a site and builds a point cloud of a3D terrain of the site including obstacle information. A voxel grid is then created from the UAV-generated point cloud, and simulation for laser scan planning is conducted to determine the stationary laser scan positions at which a mobile robot can collect data with less occluded views while capturing crucial geometric information as much as it can. Finally, optimal paths for the mobile robot to navigate among the estimated scan positions are generated. Promising test results were obtained from a real-world outdoor structural material test yard as an example of a cluttered environment. It is expected that the proposed UAV-assisted robotic approach can significantly reduce human intervention and time for data collection and processing and provide technologies to enable cluttered environments to be frequently monitored, updated, and analyzed to support timely decision-making.

Research paper thumbnail of Automation in Construction Multisensor-driven Real-time Crane Monitoring System for Blind Lift Operations: Lessons Learned from a Case Study

Automation in Construction, 2021

Crane operators are often unable to identify collision hazards due to limited visibility during b... more Crane operators are often unable to identify collision hazards due to limited visibility during blind lifts or when operating under cluttered conditions. This paper presents a multisensor-driven real-time crane monitoring system consisting of load tracking, obstacle detection, worker detection, collision warning, and 3D visualization modules. A combination of encoders, vision, and laser scanning systems is used to reconstruct a 3D workspace model of the crane environment and provide real-time spatial feedback to the operator. Field experimentation was carried out at an outdoor crane manufacturing facility under different blind lift scenarios. Results showed that the encoder-based load positioning system has a mean height estimation error within 0.33m, the vision-based load positioning system has a mean centroid error of 0.454m, and the worker detection system has a mean centroid error of 0.023m. This study also provided important findings in terms of challenges and limitations in implementing a real-world crane monitoring system.

Research paper thumbnail of Motion Planning of Mobile Robots for Autonomous Navigation on Uneven Ground Surfaces

Journal of Computing in Civil Engineering, 2021

With the increasing interest in a mobile robot for construction applications, autonomous navigati... more With the increasing interest in a mobile robot for construction applications, autonomous navigation in unstructured and uneven construction sites has become a critical challenge. To ensure the safe and robust navigation of a robot in such environments, this study develops an Optimal Obstacle-avoiding Path planner for Stable posture (OOPS) that minimizes the distance to the goal area and stabilizes the posture of the robot. In this study, a new algorithm has been developed by significantly improving the Quick-Rapidly Exploring Random Tree* (Q-RRT*) algorithm to more quickly find a feasible path and converge to the optimal path. A cost function is defined to stabilize the posture of the robot. A simulation experiment is conducted to examine the feasibility of the OOPS algorithm, and its performance is compared with those of other algorithms. The OOPS algorithm is also validated with an experiment in a real-world outdoor environment, where sloped hills exist with several obstacles. The results demonstrate that the OOPS algorithm outperforms other algorithms in terms of the time needed to find the initial solution, time to convergence on the optimal solution, and rate of success in reaching the goal. A mobile robot with the OOPS algorithm is able to start navigating sooner, and the OOPS algorithm produces paths that are closer to the optimal path in a shorter time than other 2 algorithms. The posture is more stabilized when it follows the path obtained from the OOPS algorithm in both simulation and real-world tests. Therefore, the OOPS algorithm is expected to be practically applicable for applications in uneven, highly sloped, and unstructured outdoor environments.

Research paper thumbnail of Resilient Fire Prevention and Management Strategies for Structures and Materials Stored under Urban Bridges

The collapse of an urban bridge resulting from a massive fire causes significant economic damage.... more The collapse of an urban bridge resulting from a massive fire causes significant economic damage. Preventing a recurrence of such accidents involving massive fire requires special management regulations and strategies for the fires that break out under bridges. This study introduces several strategies for the fire risk management of materials stored under bridges through the lessons learned from similar cases in South Korea. In addition, this study investigates the optimum fire management method for the space under bridges based on the best practices conducted in Korea. This study also conducted fire simulations for materials temporarily stored under bridges with different heights and volumes of the materials. Based on the results of the fire simulation, this study establishes a required distance between bridges and facilities or materials under bridges. Finally, this study suggests a guideline to establish fire management standards for the materials and facilities under bridges.

Research paper thumbnail of Deep learning-based UAV image segmentation and inpainting for generating vehicle-free orthomosaic

International Journal of Applied Earth Observation and Geoinformation, 2022

A low-altitude orthomosaic derived by an unmanned aerial vehicle (UAV) has been widely utilized f... more A low-altitude orthomosaic derived by an unmanned aerial vehicle (UAV) has been widely utilized for various purposes in large-scale infrastructure management. However, unwanted objects, such as cars and trucks, captured in the aerial images captured by the UAV have negative impacts on the quality of orthomosaic. To this end, this study presented a novel method to remove the effect of unwanted objects on UAV-generated orthomosaic. The proposed method applied a deep learning-based image segmentation and inpainting algorithm to remove the vehicles from individual UAV images before processing structure from motion (SfM), and then it resulted in generateing an orthomosaic with the inpainted UAV images. To validate the proposed method, this study conducted a case study in actual highway environment and compared the performance of the proposed method with that of another method, which directly removes and inpaints vehicles from the final orthomosaic. Through comparison tests, it is shown that the proposed method is more effective than the other. The proposed automatic vehicle-free orthomosaic generation method can contribute to creating up-to-date immersive content for transportation infrastructure management.

Research paper thumbnail of Assessment of Construction Points for Grade Control and Reference in 3D

The objective of this research is to investigate the causes of inaccuracies of grade elevations f... more The objective of this research is to investigate the causes of inaccuracies of grade elevations from current practices and identify best practices for grade control and referencing on transportation infrastructure construction projects in Georgia. A nationwide questionnaire survey was conducted with contractors, engineers, surveyors, and owners, and the following causes of the discrepancies are identified: (1) long time gap between design and construction phases; (2) poor techniques in surveying; and (3) long interval between survey points. Through an actual case study with an ongoing highway construction project, the following conclusions and recommendations are made: (1) a mandatory field survey before earth moving is recommended; (2) unmanned aerial vehicle (UAV) and Mobile Mapping System (MMS) are effective for 3D grade control in planning and site cleaning phases; (3) terrestrial laser scanning (TLS) is appropriate for 3D grade control in a subgrade grading stage; and (4) robotic total station (RTS) is recommended for the final grading stage.

Research paper thumbnail of CNN-Based Person Detection Using Infrared Images for Night-Time Intrusion Warning Systems

Sensors

Night-time surveillance is important for safety and security purposes. For this reason, several s... more Night-time surveillance is important for safety and security purposes. For this reason, several studies have attempted to automatically detect people intruding into restricted areas by using infrared cameras. However, detecting people from infrared CCTV (closed-circuit television) is challenging because they are usually installed in overhead locations and people only occupy small regions in the resulting image. Therefore, this study proposes an accurate and efficient method for detecting people in infrared CCTV images during the night-time. For this purpose, three different infrared image datasets were constructed; two obtained from an infrared CCTV installed on a public beach and another obtained from a forward looking infrared (FLIR) camera installed on a pedestrian bridge. Moreover, a convolution neural network (CNN)-based pixel-wise classifier for fine-grained person detection was implemented. The detection performance of the proposed method was compared against five conventiona...

Research paper thumbnail of Point Cloud Information Modeling: Deep Learning-based Automated Information Modeling Framework for Point Cloud Data

Journal of Construction Engineering and Management, 2021

Since point clouds have reality measurements of physical objects, they are often used to reconstr... more Since point clouds have reality measurements of physical objects, they are often used to reconstruct the as-built 3D model of building construction sites through a modeling process called Scan-to-BIM. However, the reality measurements in point cloud data, such as actual color and deformed shapes of the original objects, could disappear during the solid modeling process. In addition, the conventional Scan-to-BIM pipeline still requires significant time and manual effort for object classification and shape representation. To address these problems, this study proposes a novel information modeling framework for point clouds, called point cloud information modeling (PCIM). PCIM can automatically recognize construction objects and their properties with deep learning approaches. Furthermore, it can store information in the original point cloud data with a hierarchical structure, rather than converting it to a solid or rigid model. To validate the overall PCIM concept, this research conducted a case study with an actual building construction project. The test results demonstrate that PCIM can be an effective tool for the as-is information modeling of structures and facilities during construction.

Research paper thumbnail of Framework for Automated Registration of UAV and UGV Point Clouds Using Local Features in Images

Automation in Construction, 2019

Automatically registering 3D point clouds generated by unmanned aerial and ground vehicles (UAVs ... more Automatically registering 3D point clouds generated by unmanned aerial and ground vehicles (UAVs and UGVs) is challenging, as data is acquired at different locations with different sensors, consequently resulting in different spatial scales and occlusions. To address these problems, this study proposes a framework for the automated registration of UAV and UGV point clouds using 2 2D local feature points in the images taken from UAVs and UGVs. This study first conducted field experiments by varying the angles of the UAV camera to identify the optimal angle with which to detect sufficient points matching with the images taken by the UGV. As a result, this study identified that a combination of UAV images taken at 30˚ and 90˚ is appropriate for generating a sufficient number of matching points and attaining a reasonable level of precision. The UAV and UGV point clouds were initially scaled and registered with a transformation matrix computed from the 3D points corresponding to the 2D feature matching points. The initially aligned point clouds were subsequently adjusted by the Iterative Closest Point (ICP) algorithm, resulting in the root mean square error (RMSE) of 0.112 meters. This promising result indicates that full automation of spatial data collection and registration from a scattered environment (e.g., construction or disaster sites) by UAVs and UGVs is feasible without human intervention.

Research paper thumbnail of As-is Geometric Data Collection and 3D Visualization through the Collaboration between UAV and UGV

the 36th International Symposium on Automation and Robotics in Construction (ISARC), 2019

The characteristics of dynamic construction sites increase the difficulty of collecting the high-... more The characteristics of dynamic construction sites increase the difficulty of collecting the high-quality geometric data necessary to achieve construction management activities. This paper introduces a new autonomous framework for 3D geometric data collection in a dynamic cluttered environment using an unmanned ground vehicle (UGV) and an unmanned aerial vehicle (UAV). This method first deploys UAV to collect photo images of a site and builds a point cloud of the 3D terrain of the site, including obstacle information. A mesh grid is then created from the UAV-generated point cloud, and simulation for laser-scan planning is conducted to determine the stationary laser-scan positions at which a mobile robot can collect data with less occluded views while capturing all crucial geometric information. Finally, optimal paths for the UGV to navigate among the estimated scan positions are generated. Promising test results regarding data accuracy and collection time show that the proposed collaborative UAV-UGV approach can significantly reduce human intervention and provide technologies for enabling construction site to be frequently monitored, updated, and analyzed for timely decision-making.

Research paper thumbnail of UAV-assisted autonomous mobile robot navigation for as-is 3D data collection and registration in cluttered environments

Automation in Construction, 2019

This paper discusses enabling autonomous mobile robots to operate in unstructured terrain and env... more This paper discusses enabling autonomous mobile robots to operate in unstructured terrain and environments and gather data in a time-efficient manner. In rugged environments such as construction sites and disaster sites, spatiotemporal data is difficult to acquire since these environments are potentially hazardous to humans. Autonomous robot systems represent a reasonable solution to provide topographic and surveillance data that can assist human activity in these environments, whether for exploration, mapping, or search and rescue. However, automatically operating the mobile laser scanning robots at cluttered environments is challenging because as-is geometrical conditions of the site are difficult to comprehend from the ground level due to the blocked line-of-site views. To address the issue, this paper introduces a new framework for operating mobile robots equipped with a laser scanning system in cluttered outdoor environments with the aid of an unmanned aerial vehicle (UAV). To obtain an initial map from the current field, this method first deploys UAV to collect photographic images of a site and builds a point cloud of a3D terrain of the site including obstacle information. A voxel grid is then created from the UAV-generated point cloud, and simulation for laser scan planning is conducted to determine the stationary laser scan positions at which a mobile robot can collect data with less occluded views while capturing crucial geometric information as much as it can. Finally, optimal paths for the mobile robot to navigate among the estimated scan positions are generated. Promising test results were obtained from a real-world outdoor structural material test yard as an example of a cluttered environment. It is expected that the proposed UAV-assisted robotic approach can significantly reduce human intervention and time for data collection and processing and provide technologies to enable cluttered environments to be frequently monitored, updated, and analyzed to support timely decision-making.

Research paper thumbnail of Automation in Construction Multisensor-driven Real-time Crane Monitoring System for Blind Lift Operations: Lessons Learned from a Case Study

Automation in Construction, 2021

Crane operators are often unable to identify collision hazards due to limited visibility during b... more Crane operators are often unable to identify collision hazards due to limited visibility during blind lifts or when operating under cluttered conditions. This paper presents a multisensor-driven real-time crane monitoring system consisting of load tracking, obstacle detection, worker detection, collision warning, and 3D visualization modules. A combination of encoders, vision, and laser scanning systems is used to reconstruct a 3D workspace model of the crane environment and provide real-time spatial feedback to the operator. Field experimentation was carried out at an outdoor crane manufacturing facility under different blind lift scenarios. Results showed that the encoder-based load positioning system has a mean height estimation error within 0.33m, the vision-based load positioning system has a mean centroid error of 0.454m, and the worker detection system has a mean centroid error of 0.023m. This study also provided important findings in terms of challenges and limitations in implementing a real-world crane monitoring system.

Research paper thumbnail of Motion Planning of Mobile Robots for Autonomous Navigation on Uneven Ground Surfaces

Journal of Computing in Civil Engineering, 2021

With the increasing interest in a mobile robot for construction applications, autonomous navigati... more With the increasing interest in a mobile robot for construction applications, autonomous navigation in unstructured and uneven construction sites has become a critical challenge. To ensure the safe and robust navigation of a robot in such environments, this study develops an Optimal Obstacle-avoiding Path planner for Stable posture (OOPS) that minimizes the distance to the goal area and stabilizes the posture of the robot. In this study, a new algorithm has been developed by significantly improving the Quick-Rapidly Exploring Random Tree* (Q-RRT*) algorithm to more quickly find a feasible path and converge to the optimal path. A cost function is defined to stabilize the posture of the robot. A simulation experiment is conducted to examine the feasibility of the OOPS algorithm, and its performance is compared with those of other algorithms. The OOPS algorithm is also validated with an experiment in a real-world outdoor environment, where sloped hills exist with several obstacles. The results demonstrate that the OOPS algorithm outperforms other algorithms in terms of the time needed to find the initial solution, time to convergence on the optimal solution, and rate of success in reaching the goal. A mobile robot with the OOPS algorithm is able to start navigating sooner, and the OOPS algorithm produces paths that are closer to the optimal path in a shorter time than other 2 algorithms. The posture is more stabilized when it follows the path obtained from the OOPS algorithm in both simulation and real-world tests. Therefore, the OOPS algorithm is expected to be practically applicable for applications in uneven, highly sloped, and unstructured outdoor environments.