Automatic Landmark Detection in 2D images: A tree-based approach with multiresolution pixel features

In this paper, we propose a new generic landmark detection method for 2D images. Our solution is based on the use of ensembles of Extremely Randomized Trees combined with simple pixel-based multi-resolution features. We apply our method on a novel dataset of microscopic zebrafish images. This method was also tested on datasets of cephalometric images during the Automatic Cephalometric X-Ray Landmark Detection Challenge 2014, where we were ranked first during the first phase, and second during the second phase.

[1]  Maurício Pamplona Segundo,et al.  Automatic Face Segmentation and Facial Landmark Detection in Range Images , 2010, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[2]  Amandeep Kaur,et al.  Automatic cephalometric landmark detection using Zernike moments and template matching , 2015, Signal Image Video Process..

[3]  Pierre Geurts,et al.  Extremely randomized trees , 2006, Machine Learning.

[4]  Raphaël Marée,et al.  Automatic Localization of Interest Points in Zebrafish Images with Tree-Based Methods , 2011, PRIB.

[5]  Mandy Eberhart,et al.  Decision Forests For Computer Vision And Medical Image Analysis , 2016 .

[6]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[7]  Mariano Alcañiz Raya,et al.  Automatic Localization of Cephalometric Landmarks , 2001, J. Biomed. Informatics.

[8]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[9]  Jan Flusser,et al.  Image registration methods: a survey , 2003, Image Vis. Comput..