Automatic Generation of Training Data for Image Classification of Road Scenes

Kuhner, Tilman; Wirges, Sascha; Lauer, Martin

doi:10.1109/itsc.2019.8917089

Cited by 2 publications

(5 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [25], the average and the variance of the normalized height difference are used as features to determine which LiDAR3D points belong to the surface and those considered as curb candidates. This feature has been applied to LiDAR3D data in [17,19,22,24,33,41,50,51,55,59,62,65].…”

Section: Height Stepmentioning

confidence: 99%

“…Similarly, normals have been computed on depth images in [30] while surface normals on the 3D information obtained from a disparity map has been discussed in [65]. In [59], the normal orientation is used to verify that curb candidate points separate two horizontal planes representing the sidewalk and the road.…”

Section: Normal Orientationmentioning

confidence: 99%

“…There a variety of applications of road curb detection in intelligent transportation systems and autonomous mobile robotics, such as curbs/obstacle mapping [79], road segmentation as pre-processing for road marking classification [18,49], automatic label generation for curb detection [59], road curb detection performance toolbox [80], etc.…”

Section: Applicationsmentioning

confidence: 99%

“…This tool is intended to be method-and sensor-independent and to help improving road curb detection parameter settings. The authors of [59] automatically generate labeled datasets including curbs detected using geometrical features such as elevation and surface normals. Ring compression curb features have been used in [34] to identify the road shape using an artificial neural network classifier.…”

Section: Applicationsmentioning

confidence: 99%

See 3 more Smart Citations

Road Curb Detection: A Historical Survey

Romero

Guerrero

Romero

2021

Sensors

View full text Add to dashboard Cite

Curbs are used as physical markers to delimit roads and to redirect traffic into multiple directions (e.g., islands and roundabouts). Detection of road curbs is a fundamental task for autonomous vehicle navigation in urban environments. Since almost two decades, solutions that use various types of sensors, including vision, Light Detection and Ranging (LiDAR) sensors, among others, have emerged to address the curb detection problem. This survey elaborates on the advances of road curb detection problems, a research field that has grown over the last two decades and continues to be the ground for new theoretical and applied developments. We identify the tasks involved in the road curb detection methods and their applications on autonomous vehicle navigation and advanced driver assistance system (ADAS). Finally, we present an analysis on the similarities and differences of the wide variety of contributions.

show abstract

Section: Height Stepmentioning

confidence: 99%

Section: Normal Orientationmentioning

confidence: 99%

Section: Applicationsmentioning

confidence: 99%

Section: Applicationsmentioning

confidence: 99%

See 2 more Smart Citations

Road Curb Detection: A Historical Survey

Romero

Guerrero

Romero

2021

Sensors

View full text Add to dashboard Cite

show abstract

“…Some researchers have also begun to use other sensor modalities to automatically label training data. Kuhner et al [11] propose the use of LiDAR in driverless cars to automatically create semantic labels for the image data generated by the car's cameras. Their approach quickly annotates images of roads and curbs for use in training neural networks used to detect and navigate around these objects.…”

Section: Related Workmentioning

confidence: 99%

Automatic Generation of Machine Learning Synthetic Data Using ROS

Hart

Goodman

O'Shea

2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Data labeling is a time intensive process. As such, many data scientists use various tools to aid in the data generation and labeling process. While these tools help automate labeling, many still require user interaction throughout the process. Additionally, most target only a few network frameworks. Any researchers exploring multiple frameworks must find additional tools or write conversion scripts. This paper presents an automated tool for generating synthetic data in arbitrary network formats. It uses Robot Operating System (ROS) and Gazebo, which are common tools in the robotics community. Through ROS paradigms, it allows extensive user customization of the simulation environment and data generation process. Additionally, a plugin-like framework allows the development of arbitrary data format writers without the need to change the main body of code. Using this tool, the authors were able to generate an arbitrarily large image dataset for three unique training formats using approximately 15 minutes of user setup time and a variable amount of hands-off run time, depending on the dataset size. The source code for this data generation tool is

show abstract

Automatic Generation of Training Data for Image Classification of Road Scenes

Cited by 2 publications

References 24 publications

Road Curb Detection: A Historical Survey

Road Curb Detection: A Historical Survey

Automatic Generation of Machine Learning Synthetic Data Using ROS

Contact Info

Product

Resources

About