Road estimation with sparse 3D points from stereo data

Shinzato, Patrick Y.; Gomes, Diego; Wolf, Denis F.

doi:10.1109/itsc.2014.6957936

Cited by 22 publications

(12 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At the same time, other similar methods in KITTI-road are also compared for reference including HistonBoost, 46 BL, 4 RES3D-Stereo, 17 and BM. 47 In order to compare a learning-based method, a subpart of the method in equation (20) called ColorBoost is used as an example.…”

Section: Baselinementioning

confidence: 99%

“…In order to evaluate the performance of the proposed algorithm (ours is short for Geo þ GPR þ CRF and ours þ color for Geo þ Colorþ GPR þ CRF), a set of comparisons on KITTI-road data set are taken including HistonBoost, 46 BL, 4 RES3D-Stereo, 17 BM, 47 and ColorBoost.…”

Section: Multifeatþcrf Versus Multifeatþgprþcrfmentioning

confidence: 99%

“…Note that the results are evaluated on the latest version of ground truth, while the results in HistonBoost, 46 RES3D-Stereo, 17 and BM 47 are from the corresponding literatures. Some of them may be evaluated on old version criteria but they will not vary dramatically.…”

Section: Multifeatþcrf Versus Multifeatþgprþcrfmentioning

confidence: 99%

“…Mendes et al 16 modified this algorithm in a dynamically sized sliding windows framework on a GPU to reduce the computation as well. Some varied approaches are proposed, 17,18 which use similar geometry principles to get a probability representation and apply Delaunay triangulation to generate the graph model for free space detection.…”

Section: Related Workmentioning

confidence: 99%

See 3 more Smart Citations

Gaussian process regression-based robust free space detection for autonomous vehicle by 3-D point cloud and 2-D appearance information fusion

Xiao

Dai

et al. 2017

International Journal of Advanced Robotic Systems

View full text Add to dashboard Cite

Free space detection is crucial to autonomous vehicles while existing works are not entirely satisfactory. As cameras have many advantages on environment perception, a stereo vision-based robust free space detection method is proposed which mainly depends on geometry information and Gaussian process regression. In this work, in order to improve the performance by exploiting multiple source information, we apply Bayesian framework and conditional random field inference to fuse the multimodal information including 2-D image and 3-D point geometric information. Particularly, a Bayesian framework is used for multiple feature fusion to provide a normalized and flexible output. Gaussian process regression is used to automatically and incrementally regress the data, resulting enhanced performance. Finally, conditional random field with color and geometry constrains is applied to make the result more robust. In order to evaluate the proposed method, quantitative experiments on popular KITTI-road data set and qualitative experiments on our own campus data set are tested. The results show satisfactory and inspiring performance compared to the outstanding works and even are competitive to some relevant Lidar-based methods.

show abstract

Section: Baselinementioning

confidence: 99%

Section: Multifeatþcrf Versus Multifeatþgprþcrfmentioning

confidence: 99%

Section: Multifeatþcrf Versus Multifeatþgprþcrfmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Gaussian process regression-based robust free space detection for autonomous vehicle by 3-D point cloud and 2-D appearance information fusion

Xiao

Dai

et al. 2017

International Journal of Advanced Robotic Systems

View full text Add to dashboard Cite

show abstract

“…all methods already has some kind of previous information (model of features) about the environment. A brief comparison between our current method (GRES3D) and our previous methods (RES3D-Velo (Shinzato et al, 2014a) and RES3D-Stereo (Shinzato et al, 2014b)) shows a significant improvement in our runtime which is 6× faster for the 3D-LIDAR approach and almost 7× faster for the stereo approach. Such an improvement was achieved because our current approach does not require intermediate filters and does not compute disparity, neither classifies all the pixels, but only points selected to compose the graph.…”

Section: Online Kitti-road Benchmarkmentioning

confidence: 80%

Estimação de obstáculos e área de pista com pontos 3D esparsos

Shinzato¹

View full text Add to dashboard Cite

Worldwide, an estimated 1.2 million lives are lost in road crashes each year and Advanced Driver Assistance Systems (ADAS) and Self-driving cars promise to reduce this number. Among the various issues to complete this technology, perception systems are still an unsolved issues. Normally two of them, obstacle detection and road detection, make use of sophisticated algorithms such as supervised machine learning methods which can perform with impressive results if it was trained with good datasets. Since it is a complex and an expensive job to create and maintain databases of scenarios from the entire world, adaptive and/or self-supervised methods are good candidates for detection systems in the near future. Due that, this thesis present a method to estimate obstacles and estimate the road terrain using low cost sensors (stereo camera), avoiding supervised machine learning techniques and the most common assumptions used by works presented in literature. These methods were compared with 3D-LIDAR approaches achieving similar results and thus it can be used as a pre-processing step to improve or allow adaptive methods with machine learning systems.i ResumoDe acordo com a Organização Mundial da Saúde, cerca de 1,2 milhões de pessoas no mundo morrem em acidentes de trânsito. Sistemas de assistência ao motorista e veículos autônomos podem diminuir o número de acidentes. Dentre as várias demandas existentes para viabilizar essa tecnologia, sistemas computacionais de percepção ainda permanecem sem uma solução definitiva. Dois deles, detecção de obstáculos e de via navegável, normalmente fazem uso de algoritmos sofisticados como técnicas de aprendizado supervisionado, que mostram resultados impressionantes quando treinados com bases de dados bem definidas e diversificadas. Entretanto, construir, manter e atualizar uma base de dados com exemplos de vários lugares do mundo e em diversas situações é trabalhoso e complexo. Assim, métodos adaptativos e auto-supervisionados mostram-se como boas alternativas para sistemas de detecção do futuro próximo. Neste contexto, esta tese apresenta um método para estimar obstáculos e via navegável através de sensores de baixo custo (câmeras estereoscópicas), sem o uso de técnicas de aprendizado de máquina e de diversas suposições normalmente utilizadas por trabalhos ja disponíveis na literatura. Esses métodos utilizando sensor estereoscópico foram comparados fazendo uso de sensores do tipo 3D-LIDAR e mostraram resultados semelhantes. Este sistema poderá ser usado como uma fase pré-processamento de dados para melhorar ou viabilizar métodos adaptativos de aprendizado.iii Acknowledgements AgradecimentosEis que aqui se encontra o resultado de alguns anos de trabalho. Um resultado que significa o fim e o início de uma etapa da vida de todo aquele que procura contribuir de alguma forma para a ciência e benefício da sociedade. Por isso, não poderia deixar de colocar neste documento, o meu profundo agradecimento a todos que de alguma forma contribuíram para a execução deste trabalho. Neste s...

show abstract