Publication detail

Evolutionary Improved Object Detector for Ultrasound Images

MAŠEK, J. BURGET, R. KARÁSEK, J. UHER, V. GÜNEY, S.

Original Title

Evolutionary Improved Object Detector for Ultrasound Images

English Title

Evolutionary Improved Object Detector for Ultrasound Images

Type

conference paper

Language

en

Original Abstract

Object detection in ultrasound images is difficult problem mainly because of relatively low signal–to–noise ratio. This paper deals with object detection in the noisy ultrasound images using modified version of Viola–Jones object detector. The method describes detection of carotid artery longitudinal section in ultrasound B–mode images. The detector is primarily trained by AdaBoost algorithm and uses a cascade of Haar–like features as a classifier. The main contribution of this paper is a method for detection of carotid artery longitudinal section. This method creates cascade of classifiers automatically using genetic algorithms. We also created post–processing method that marks position of artery in the image. The proposed method was released as open–source software. Resulting detector achieved accuracy 96.29%. When compared to SVM classification enlarged with RANSAC (RANdom SAmple Consensus) method that was used for detection of carotid artery longitudinal section, works our method real–time.

English abstract

Object detection in ultrasound images is difficult problem mainly because of relatively low signal–to–noise ratio. This paper deals with object detection in the noisy ultrasound images using modified version of Viola–Jones object detector. The method describes detection of carotid artery longitudinal section in ultrasound B–mode images. The detector is primarily trained by AdaBoost algorithm and uses a cascade of Haar–like features as a classifier. The main contribution of this paper is a method for detection of carotid artery longitudinal section. This method creates cascade of classifiers automatically using genetic algorithms. We also created post–processing method that marks position of artery in the image. The proposed method was released as open–source software. Resulting detector achieved accuracy 96.29%. When compared to SVM classification enlarged with RANSAC (RANdom SAmple Consensus) method that was used for detection of carotid artery longitudinal section, works our method real–time.

Keywords

carotid artery, genetic algorithms, ultrasound, object detection, Viola–Jones detector.

RIV year

2013

Released

02.07.2013

ISBN

978-1-4799-0402-0

Book

36th International Conference on Telecommunications and Signal processing

Pages from

586

Pages to

590

Pages count

5

BibTex


@inproceedings{BUT100842,
  author="Jan {Mašek} and Radim {Burget} and Jan {Karásek} and Václav {Uher} and Selda {Güney}",
  title="Evolutionary Improved Object Detector for Ultrasound Images",
  annote="Object detection in ultrasound images is difficult problem mainly because of relatively low signal–to–noise ratio. This paper deals with object detection in the noisy ultrasound images using modified version of Viola–Jones object detector. The method describes detection of carotid artery longitudinal section in ultrasound B–mode images. The detector is primarily trained by AdaBoost algorithm and uses a cascade of Haar–like features as a classifier. The main contribution of this paper is a method for detection of carotid artery longitudinal section. This method creates cascade of classifiers automatically using genetic algorithms. We also created post–processing method that
marks position of artery in the image. The proposed method was released as open–source software. Resulting detector achieved accuracy 96.29%. When compared to SVM classification enlarged with RANSAC (RANdom SAmple Consensus) method that was used for detection of carotid artery longitudinal section, works our method real–time.",
  booktitle="36th International Conference on Telecommunications and Signal processing",
  chapter="100842",
  doi="10.1109/TSP.2013.6614002",
  howpublished="print",
  year="2013",
  month="july",
  pages="586--590",
  type="conference paper"
}