An Adaptive Feature-fusion Method for Object Matching over Non-overlapped Scenes

Liu, Huanxi; Lv, Xiaowei; Zhu, Tianhong; Li, Xiong

doi:10.1007/s11265-013-0806-7

Cited by 3 publications

(4 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To further validate the effectiveness of our approach, we compare our results with the latest person re-identification approaches in Refs. [1,2,6,12,13] on two multi-shot datasets: ETHZ and MCT. The results are shown in Fig.8.…”

Section: Comparison With Current Methodsmentioning

confidence: 99%

“…The interval difference degree [18] is applied to evaluate the performance of the similarity measure function, and it is expressed as Eq. (1).…”

Section: Significant Difference Distancementioning

confidence: 99%

“…Person re-identification plays an important role in cross-camera tracking [1] , and it has drawn a lot of attention in recent years [2] . Person re-identification is used to match pedestrians observed from different camera views by using some appearance features [3,4] .…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Person Re‐identification Across Multiple Non‐overlapping Cameras by Grouping Similarity Comparison Model

et al. 2017

View full text Add to dashboard Cite

Section: Comparison With Current Methodsmentioning

confidence: 99%

“…The interval difference degree [18] is applied to evaluate the performance of the similarity measure function, and it is expressed as Eq. (1).…”

Section: Significant Difference Distancementioning

confidence: 99%

See 1 more Smart Citation

Person Re‐identification Across Multiple Non‐overlapping Cameras by Grouping Similarity Comparison Model

et al. 2017

View full text Add to dashboard Cite

“…So finding strong image features to realize robust object matching between different cameras is studied. A method based region SIFT feature extraction has been proposed to solve the object matching problem with shooting angle limitation [3].Zhou Q integrated several image features such as color feature, contour feature to realized more robust object matching with computational complex cost [4].Guo Y used several image features to realize vehicle tracking in multi-camera system [5].Huanxi Liu proposed an adaptive feature integrated method for object matching including color histogram, UV chrominance, spectrum and SIFT [6].…”

Section: Introductionmentioning

confidence: 99%

Research on the Collaborative Tracking system for Moving Target in multi-camera surveillance system based on space-time video fusion

Zhao¹,

Yu²,

Yu³

et al. 2015

Proceedings of the 4th International Conference on Mechatronics, Materials, Chemistry and Computer Engineering 2015

View full text Add to dashboard Cite

With the increasing demand of social security, intelligent moving target tracking technology play a more important role in video surveillance system. However, current moving target tracking technology works well for single camera situation but always lost targets in multi-camera environments due to the difficulty of object matching. In this paper, we proposed a new collaborative tracking system for moving target based on space-time video fusion .The idea of this system is firstly to register and fuse real-time surveillance video to 3D virtual model with situation awareness, then the space relativity between different cameras and the time relativity of tracking target moving across multi-cameras are extracted. With the help of this space-time information, we can accurately match the moving targets between consecutive cameras and coordinate adjacent cameras to collaboratively track the target by using moving target tracking for single camera. Test results show that our approach can realize continuous tracking for moving objects between multi-cameras with good robustness.

show abstract

Specific object re-identification across non-overlapping camera views in traffic accidents

Tan

Huang

Zhang

et al. 2016

Transportation Planning and Technology

View full text Add to dashboard Cite

An Adaptive Feature-fusion Method for Object Matching over Non-overlapped Scenes

Cited by 3 publications

References 8 publications

Person Re‐identification Across Multiple Non‐overlapping Cameras by Grouping Similarity Comparison Model

Person Re‐identification Across Multiple Non‐overlapping Cameras by Grouping Similarity Comparison Model

Research on the Collaborative Tracking system for Moving Target in multi-camera surveillance system based on space-time video fusion

Specific object re-identification across non-overlapping camera views in traffic accidents

Contact Info

Product

Resources

About