Continuous Machine Learning in Computer Vision - Tracking with Adaptive Class Models

Stolkin, Rustam

doi:10.5772/4938

Cited by 3 publications

(2 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The detection component seeks to identify objects of interest in the FOV. To date, the principal methods involve background modeling, target modeling and matching, and machine learning based methods [62][63][64][65][66][67][68][69]. Once detected, tracking algorithms can then be applied to obtain an estimation of the motion of the target in space and time.…”

Section: Computer Vision For Construction Operations Analysismentioning

confidence: 99%

Status quo and open challenges in vision-based sensing and tracking of temporary resources on infrastructure construction sites

Teizer¹

2015

Advanced Engineering Informatics

149

View full text Add to dashboard Cite

Section: Computer Vision For Construction Operations Analysismentioning

confidence: 99%

Status quo and open challenges in vision-based sensing and tracking of temporary resources on infrastructure construction sites

Teizer¹

2015

Advanced Engineering Informatics

149

View full text Add to dashboard Cite

“…the strength and direction of the EM field) and sends this information back to a computer. The computer triangulates the location (position and orientation) of the detector relative to the three EM fields (Kuipers, 1980; Polhemus Manual, 2017; Raab et al, 1979; Stolkin, 2014).…”

Section: Introductionmentioning

confidence: 99%

Automated identification of distinct events in an assembly process using tracking of body postures

Balaraman

Chakrabarti

Gurumoorthy

et al. 2017

Concurrent Engineering

View full text Add to dashboard Cite

An ergonomic assembly process is safer for operators and is more efficient in quality, time, cost and productivity. An assembly process is carried out in a series of distinct events. Ergonomic assessment of an assembly process therefore involves identification of the distinct events within the process and assessment of difficulty in carrying out each event, so that the process can be improved in an event-specific manner. So far, such assessment of assembly is carried out using video by identifying key frames manually. Manual identification of key frames consumes more time. To resolve this drawback, a novel approach has been proposed for tracking the body segments using electromagnetic trackers. It is done automatically in real time. Then, a data smoothening method is used for analysing automatically the tracked data to identify distinct events in an assembly process. An experiment in a laboratory setting is used in this study to test the following hypothesis: 'An event is characterized by gross movements at its beginning and its end'. This hypothesis encapsulates the essence of the signature in postural data, which is used by the proposed method for identifying distinct events within an assembly; the tracked data were used to identify this signature automatically from the trackers and the tracked data set with a threshold of movement for each body segment; using this threshold, gross motions and thereby starts and ends of distinct events can be identified. The method of identifying events was based on a reach experiment; thus, the variation of torso angle was studied in this work. The result of the study indicates that this method can be applied easily in detecting events in an assembly. Performance of the proposed method is compared with the traditional method and it is observed that the proposed approach outperforms traditional method in terms of time and accuracy.

show abstract

Robust visual tracking based on adaptive depth-color-cue integration using range sensor

Wang

Liu

2012

2012 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI)

View full text Add to dashboard Cite

Continuous Machine Learning in Computer Vision - Tracking with Adaptive Class Models

Cited by 3 publications

References 19 publications

Status quo and open challenges in vision-based sensing and tracking of temporary resources on infrastructure construction sites

Status quo and open challenges in vision-based sensing and tracking of temporary resources on infrastructure construction sites

Automated identification of distinct events in an assembly process using tracking of body postures

Robust visual tracking based on adaptive depth-color-cue integration using range sensor

Contact Info

Product

Resources

About