Optimal estimation of line segments in noisy lidar data

Kapp, A.; Gröll, Lutz

doi:10.1016/j.sigpro.2005.10.016

Cited by 8 publications

(12 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Moveover, the model parameters can be identified on the basis of a single scan frame. This paper adds to the authors' examination of statistical errors on the estimation of straight lines in noisy lidar data [13].…”

Section: Discussionmentioning

confidence: 98%

See 1 more Smart Citation

Effect of Fast Motion on Range Images Acquired by Lidar Scanners for Automotive Applications

Gröll

Kapp

2007

IEEE Trans. Signal Process.

Self Cite

View full text Add to dashboard Cite

Abstract-The scanning process of lidar scanners is extended over time. Hence, relative motion between sensor and object during the scanning leads to a motion-dependent blur, here called motion-scan effect. Due to this effect the lidar's distance map is deteriorated, which has a negative impact on the subsequent signal processing. This paper quantifies this effect and compares it to statistical errors. In order to compensate for the negative impact, motion models for a simple car model are formulated, and the model parameters are checked for identifiability. Identification of the parameters allows to compensate for the effect. Thus the accuracy and the control mode can be improved, which is particularly important for safety-critical applications, e.g., autonomous vehicles.

show abstract

Section: Discussionmentioning

confidence: 98%

“…The angle measurement is assumed to be noise-free while the measurement noise of the distance is modeled as independent Gaussian with standard deviation m. Details to the statistical examination can be found in [13].…”

Section: ) Comparison With Statistical Errorsmentioning

confidence: 99%

Effect of Fast Motion on Range Images Acquired by Lidar Scanners for Automotive Applications

Gröll

Kapp

2007

IEEE Trans. Signal Process.

Self Cite

View full text Add to dashboard Cite

show abstract

“…In industrial applications, there are a lot of noises in these images. If traditional edge detection methods are used to detect edges in all area of the images in our application, not only a lot of time caused by calculation would be needed but also the precision of the edge detection would be much worse since traditional edge detection methods usually have to select threshold which is much sensitive to noise [4,5] . Theoretical analysis and direct image vision prove there do be two bright stripes in an image though there are noises.…”

Section: Image Processing Of Cross Laser Imagesmentioning

confidence: 99%

Straightness inspecting system for long shaft based on cross laser

2010

View full text Add to dashboard Cite

Based on the principle that the distance from the axle of an inspected long shaft to the two laser planes is unique, the method used image processing to get the coordinate of the cross point and the angle of the horizontal line of the laser image, and calculated the straightness parameters with distance position information along the long shaft. The image processing of the cross laser was carried out by a local maximum intensity. With a camera (640×480 pixels and 0.0973mm/pixel as resolution of target), 50 samples were taken at each distance of 1m, 9m and 18m. The standard deviation was 0.00937mm, 0.0342mm and 0.0860mm for polar radius, 0.843 ＇ , 1.26 ＇ and 1.57 ＇ for angle, respectively. Experimental results indicate that the straightness inspection system for long shaft is simple in structure, easy to operate and precise in inspection results.

show abstract

“…Dies ist zum einen der einfache Least-Squares-Schätzer, welcher das Messrauschen jedoch nur unzureichend einbezieht. Zwei neuere Ansätze nach dem Maximum-Likelihood-Prinzip berücksichtigen zwar das Messrauschen in der radialen Komponente, nehmen einen Fehler in der Winkelemessung jedoch entweder nicht an [3], oder verwenden Abschätzungen für kleine Winkelfehler [5]. Der in dieser Arbeit vorgeschlagene Schätzer stellt eine Erweiterung des letztgenannten Verfahrens um eine Korrektur der Linearisierungsfehler dar.…”

unclassified

Optimale Geradenschätzung für Time-of-Flight Umfeldsensoren

Stellet

Kutzera²

2011

Tm - Technisches Messen

View full text Add to dashboard Cite

Zusammenfassung Dieser Artikel stellt einen verbesserten Algorithmus zur Geradenschätzung für Umfeldsensoren vor. Eine optimale Schätzung ist nur möglich, wenn die stochastischen Eigenschaften der Messfehler berücksichtigt werden. Lidar-und Ultraschallsensoren erfassen die radiale Entfernung r und den Azimutwinkel ϕ. Deshalb ist es sinnvoll, unkorrelierte Messfehler in Polarkoordinaten anzunehmen. Schätzverfahren, die die daraus folgende Korrelation in kartesischen Koordinaten nicht berücksichtigen, können daher einen systematischen Schätzfehler aufweisen. Zwei neuere Ansätze zur Maximum-Likelihood-Schätzung der Geradenparameter werden hier vorgestellt und es wird die Verbesserung eines der Verfahren vorgeschlagen.Summary This paper presents an improved algorithm for estimating line segments in noisy range data. An estimator's performance depends on correctly taking the measurement errors into account. Most sensors measure range and angle independently and thus it is a sound assumption to model the measurement error by independent noise processes in radius r and azimuth angle ϕ. However, this leads to correlated errors in Cartesian coordinates. Simple estimators do not consider this and therefore yield poor results. Two more recent approaches based on Maximum-Likelihood-Estimation are discussed in this paper and an improved formulation is presented.

show abstract

Optimal estimation of line segments in noisy lidar data

Cited by 8 publications

References 11 publications

Effect of Fast Motion on Range Images Acquired by Lidar Scanners for Automotive Applications

Effect of Fast Motion on Range Images Acquired by Lidar Scanners for Automotive Applications

Straightness inspecting system for long shaft based on cross laser

Optimale Geradenschätzung für Time-of-Flight Umfeldsensoren

Contact Info

Product

Resources

About