Image Segmentation Based on Velocity Feature Vector for Moving Target Extraction

Image segmentation of moving target in complex background has a good application prospect. However, many factors cause the gray level and content of the image to change between adjacent two frames. Based on the research and analysis of various moving target segmentation algorithms, this paper propos...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:IEEE sensors journal Ročník 20; číslo 20; s. 11983 - 11991
Hlavný autor: Hu, Yunfei
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: New York IEEE 15.10.2020
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Predmet:
Algorithms
Contours
Dynamic contour
Feature extraction
Heuristic algorithms
Image segmentation
Mathematical analysis
Motion segmentation
moving target
Moving targets
Occlusion
Sensors
Target recognition
Target tracking
velocity measurement
Video surveillance
ISSN:1530-437X, 1558-1748
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Image segmentation of moving target in complex background has a good application prospect. However, many factors cause the gray level and content of the image to change between adjacent two frames. Based on the research and analysis of various moving target segmentation algorithms, this paper proposes a new dynamic segmentation algorithm based on dynamic contours based on the principle of simple calculation, fast speed and accurate extraction of moving targets. Analysis and experiments prove that the algorithm needs to adjust less parameters and has strong anti-interference ability, which can effectively eliminate the occlusion of moving targets between multiple frames. Moreover, for complex backgrounds and non-overlapping multi-target motion situations, the outer contours of individual moving targets can be accurately located. In addition, the algorithm has a potential parallel mechanism and is easy to implement real-time moving image segmentation.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1530-437X
1558-1748
DOI:10.1109/JSEN.2020.2974314