Real-time monocular human height estimation using bimodal background subtraction

Wangsiripitak, Somkiat; Saelao, Wongsatorn

doi:10.1109/tencon.2017.8227847

Cited by 3 publications

(5 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In our approach, we estimate human height in each frame using an approximated ground plane, a ray from camera center to human foot bottom, a normal vector at that foot, and a ray from camera center to human head top (similar to the procedure reported in [4]), hereafter referred to as in-frame height. These in-frame heights are improved by machine learning to obtain a final height that is close to the real height.…”

Section: Methodsmentioning

confidence: 99%

“…Several research works have focused on utilizing human height information to identify a person. Jeges et al [4] proposed to obtain the height of people by calibrating the camera. They reported that estimating a human height from multiple views is more robust than doing so from a single view.…”

Section: Related Workmentioning

confidence: 99%

“…There are various ways to extract human height feature. The method reported in [4] used a calibrated camera and an estimated ground plane to estimate in-frame height. The human in the clip was detected by using background subtraction.…”

Section: Faculty Of Information Technologymentioning

confidence: 99%

“…The bottom and top edges were used as 2D positions of the bottom of the foot and the top of the head respectively-their 3D positions were determined from the intersection of the projection ray, the ground plane, and the normal vector from the foot position. The automated localization of points at the top of the head and the bottom of foot in [4], however, caused a problem-the captured data often contained outliers that led to mis-localization of those points, hence an incorrect inframe height estimate. The proposed approach attempts to alleviate this problem and to test the feasibility of applying machine learning to estimation of the final height.…”

Section: Faculty Of Information Technologymentioning

confidence: 99%

“…The proposed approach attempts to alleviate this problem and to test the feasibility of applying machine learning to estimation of the final height. Unlike the approach in [4], the proposed approach manually marks points at the top of head and the bottom of the foot to avoid that problem. Fig.…”

Section: Faculty Of Information Technologymentioning

confidence: 99%

See 4 more Smart Citations

Human Height Estimation Using Visual Geometry and Feature Learning

Dokthurian

Rattanapitak

Wangsiripitak

2021

2021 IEEE International Symposium on Circuits and Systems (ISCAS)

Self Cite

View full text Add to dashboard Cite

Many existing video surveillance systems use human characteristics like face, height, and gait to identify a person. This paper proposes a human height estimation approach using visual geometry and feature learning that makes an estimate from a video clip of a person. An experiment was conducted to evaluate the performance of the approach. The approach achieved an average percentage final height estimate of 100.59 % (actual height = 100%), better than a previously reported estimate of 98.8% in the literature achieved by another approach. A successful further development of this approach would directly benefit forensic science investigators.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Faculty Of Information Technologymentioning

confidence: 99%

Section: Faculty Of Information Technologymentioning

confidence: 99%

Section: Faculty Of Information Technologymentioning

confidence: 99%

See 3 more Smart Citations