CNN-based pre-processing and multi-frame-based view transformation for fisheye camera-based AVM system

Choi, Dong Yoon; Choi, Ji Hoon; Choi, Jin Wook; Song, Byung Cheol

doi:10.1109/icip.2017.8297048

Cited by 3 publications

(2 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These regions are critical, as they gather a big part of the scene information. To deal with this problem, in [13], a CNN-based preprocessing stage and a multi-frame-based view transformation were proposed and applied in an Around View Monitor system (AVM). However, this approach uses separated CNN frameworks for image enhancement and up-scaling and hole filling method can be improved.…”

Section: Related Workmentioning

confidence: 99%

Real-Time Semantic Segmentation for Fisheye Urban Driving Images Based on ERFNet

Sáez

Bergasa

López-Guillén

et al. 2019

Sensors

View full text Add to dashboard Cite

The interest in fisheye cameras has recently risen in the autonomous vehicles field, as they are able to reduce the complexity of perception systems while improving the management of dangerous driving situations. However, the strong distortion inherent to these cameras makes the usage of conventional computer vision algorithms difficult and has prevented the development of these devices. This paper presents a methodology that provides real-time semantic segmentation on fisheye cameras leveraging only synthetic images. Furthermore, we propose some Convolutional Neural Networks(CNN) architectures based on Efficient Residual Factorized Network(ERFNet) that demonstrate notable skills handling distortion and a new training strategy that improves the segmentation on the image borders. Our proposals are compared to similar state-of-the-art works showing an outstanding performance and tested in an unknown real world scenario using a fisheye camera integrated in an open-source autonomous electric car, showing a high domain adaptation capability.

show abstract

Section: Related Workmentioning

confidence: 99%

Real-Time Semantic Segmentation for Fisheye Urban Driving Images Based on ERFNet

Sáez

Bergasa

López-Guillén

et al. 2019

Sensors

View full text Add to dashboard Cite

show abstract

“…Estimating the pose of a camera attached to an autonomous vehicle is a method for outputting 3D data. In this paper, we propose a network that predicts an LUT [1] that transforms a camera image plane to a BEV [2] plane, and we aim to estimate the pose of the camera through this.…”

Section: Introductionmentioning

confidence: 99%

Birds Eye View Look-Up Table Estimation with Semantic Segmentation

2021

View full text Add to dashboard Cite

In this work, a study was carried out to estimate a look-up table (LUT) that converts a camera image plane to a birds eye view (BEV) plane using a single camera. The traditional camera pose estimation fields require high costs in researching and manufacturing autonomous vehicles for the future and may require pre-configured infra. This paper proposes an autonomous vehicle driving camera calibration system that is low cost and utilizes low infra. A network that outputs an image in the form of an LUT that converts the image into a BEV by estimating the camera pose under urban road driving conditions using a single camera was studied. We propose a network that predicts human-like poses from a single image. We collected synthetic data using a simulator, made BEV and LUT as ground truth, and utilized the proposed network and ground truth to train pose estimation function. In the progress, it predicts the pose by deciphering the semantic segmentation feature and increases its performance by attaching a layer that handles the overall direction of the network. The network outputs camera angle (roll/pitch/yaw) on the 3D coordinate system so that the user can monitor learning. Since the network's output is a LUT, there is no need for additional calculation, and real-time performance is improved.

show abstract

A New Approach to Use Modern Object Detection Methods More Efficiently on CCTV Systems

Can

Kantarci

Ünal

2021

2021 6th International Conference on Computer Science and Engineering (UBMK)

View full text Add to dashboard Cite

CNN-based pre-processing and multi-frame-based view transformation for fisheye camera-based AVM system

Cited by 3 publications

References 13 publications

Real-Time Semantic Segmentation for Fisheye Urban Driving Images Based on ERFNet

Real-Time Semantic Segmentation for Fisheye Urban Driving Images Based on ERFNet

Birds Eye View Look-Up Table Estimation with Semantic Segmentation

A New Approach to Use Modern Object Detection Methods More Efficiently on CCTV Systems

Contact Info

Product

Resources

About