Probabilistic Multi-class Scene Flow Segmentation for Traffic Scenes

Barth, Alexander; Siegemund, Jan; Meißner, Annemarie; Franke, Uwe; Förstner, Wolfgang

doi:10.1007/978-3-642-15986-2_51

Cited by 10 publications

(10 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Segmentation Experiment. In [3] a method for segmentation and tracking of independently moving objects (IMO) is presented. The segmentation is based on evaluating probabilities at pixels whether they are in motion in real world coordinates or static, based on using scene flow and ego-motion information.…”

Section: Experiments and Resultsmentioning

confidence: 99%

“…The segmentation is based on evaluating probabilities at pixels whether they are in motion in real world coordinates or static, based on using scene flow and ego-motion information. Although this approach is not comparable to the approach presented here, we use the ground truth image provided by [3] (publicly available in Set 7 of [5]) as indication for the quality of the segmentation part of our algorithm. After incorporating ego-motion analysis into our framework (at a later stage), segmented objects can be labelled 'static' or 'in motion', as in a scene flow approach.…”

Section: Experiments and Resultsmentioning

confidence: 99%

“…Results and Discussion. Figure 6 shows resulting stereo maps of two traffic scenes provided by the study [3] as generated by our SGM implementation with subsequent mode filtering and ground manifold removal. On the right of both maps, contours are shown that correspond to moving objects within the scene.…”

Section: Experiments and Resultsmentioning

confidence: 99%

See 2 more Smart Citations

Mid-level Segmentation and Segment Tracking for Long-Range Stereo Analysis

Hermann

Börner

Klette

2011

Advances in Image and Video Technology

View full text Add to dashboard Cite

This paper presents a novel way of combining dense stereo and motion analysis for the purpose of mid-level scene segmentation and object tracking. The input is video data that addresses long-range stereo analysis, as typical when recording traffic scenes from a mobile platform. The task is to identify shapes of traffic-relevant objects without aiming at object classification at the considered stage. We analyse disparity dynamics in recorded scenes for solving this task. Statistical shape models are generated over subsequent frames. Shape correspondences are established by using a similarity measure based on set theory. The motion of detected shapes (frame to frame) is compensated by using a dense motion field as produced by a real-time optical flow algorithm. Experimental results show the quality of the proposed method which is fairly simple to implement. The first author thanks the German Academic Exchange Service (DAAD) for financial support.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

Section: Experiments and Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Mid-level Segmentation and Segment Tracking for Long-Range Stereo Analysis

Hermann

Börner

Klette

2011

Advances in Image and Video Technology

View full text Add to dashboard Cite

show abstract

“…Using prior knowledge of the potential floor B in the scene and pixel-level information then each homogeneous region is assigned to either ground floor GF, one of the sitting places ST or static background BG. The global topological order of the classes such as resting places are above ground and background is also above ground or at same level as resting places, is locally integrated in to a CRF [35,43] by ordering constraints. Fig.…”

Section: Joint Conditional Segmentationmentioning

confidence: 99%

Estimating layout of cluttered indoor scenes using trajectory-based priors

Shoaib

Yang

Rosenhahn

et al. 2014

Image and Vision Computing

View full text Add to dashboard Cite

“…The authors of [30] use graph cut to identify moving objects based on a probabilistic threshold. In [1], Barth et al present a probabilistic multi-class traffic scene segmentation approach purely based on dense depth and motion information. This work focuses on the latter approach, which it expands in various aspects.…”

Section: Related Workmentioning

confidence: 99%

Stixmentation - Probabilistic Stixel based Traffic Scene Labeling

Erbs¹,

Schwarz²,

Franke³

2012

Procedings of the British Machine Vision Conference 2012

Self Cite

View full text Add to dashboard Cite

The detection of moving objects like vehicles, pedestrians or bicycles from a mobile platform is one of the most challenging and most important tasks for driver assistance and safety systems. For this purpose, we present a multi-class traffic scene segmentation approach based on the Dynamic Stixel World, an efficient super-pixel object representation. In this approach, each Stixel is assigned either to a quantized maneuver motion class like oncoming, or left-moving or to static background. The formulation integrates multiple 3D and motion features as well as spatio-temporal prior knowledge in a probabilistic conditional random field (CRF) framework. The real-time capable method is evaluated quantitatively in various challenging, cluttered urban traffic scenes. The experimental results yield highly accurate segmentation of urban traffic scenarios without the need for any manual parameter adjustments.

show abstract

Probabilistic Multi-class Scene Flow Segmentation for Traffic Scenes

Cited by 10 publications

References 10 publications

Mid-level Segmentation and Segment Tracking for Long-Range Stereo Analysis

Mid-level Segmentation and Segment Tracking for Long-Range Stereo Analysis

Estimating layout of cluttered indoor scenes using trajectory-based priors

Stixmentation - Probabilistic Stixel based Traffic Scene Labeling

Contact Info

Product

Resources

About