A novel feature descriptor based on microscopy image statistics. Bayramoglu, N., Kannala, J., Akerfelt, M., Kaakinen, M., Eklund, L., Nees, M., & Heikkilä, J. In Proceedings - International Conference on Image Processing, ICIP, pages 2695-2699, 2015. IEEE. abstract bibtex © 2015 IEEE. In this paper, we propose a novel feature description algorithm based on image statistics. The pipeline first performs independent component analysis on training image patches to obtain basis vectors (filters) for a lower dimensional representation. Then for a given image, a set of filter responses at each pixel is computed. Finally, a histogram representation, which considers the signs and magnitudes of the responses as well as the number of filters, is applied on local image patches. We propose to apply this idea to a microscopy image pixel identification system based on a learning framework. Experimental results show that the proposed algorithm performs better than the state-of-the-art descriptors in biomedical images of different microscopy modalities.
@inProceedings{
title = {A novel feature descriptor based on microscopy image statistics},
type = {inProceedings},
year = {2015},
identifiers = {[object Object]},
keywords = {cell co-culture,cell detection,electron microscopy,local image descriptor,mitochondria,phase contrast imaging,pixel labeling,tumor},
pages = {2695-2699},
publisher = {IEEE},
id = {06f66f46-06e4-3fb5-b62e-afaa9ebfaffe},
created = {2019-09-15T16:34:27.675Z},
file_attached = {false},
profile_id = {bddcf02d-403b-3b06-9def-6d15cc293e20},
last_modified = {2019-09-19T17:47:33.125Z},
read = {false},
starred = {false},
authored = {true},
confirmed = {true},
hidden = {false},
source_type = {CONF},
private_publication = {false},
abstract = {© 2015 IEEE. In this paper, we propose a novel feature description algorithm based on image statistics. The pipeline first performs independent component analysis on training image patches to obtain basis vectors (filters) for a lower dimensional representation. Then for a given image, a set of filter responses at each pixel is computed. Finally, a histogram representation, which considers the signs and magnitudes of the responses as well as the number of filters, is applied on local image patches. We propose to apply this idea to a microscopy image pixel identification system based on a learning framework. Experimental results show that the proposed algorithm performs better than the state-of-the-art descriptors in biomedical images of different microscopy modalities.},
bibtype = {inProceedings},
author = {Bayramoglu, Neslihan and Kannala, Juho and Akerfelt, Malin and Kaakinen, Mika and Eklund, Lauri and Nees, Matthias and Heikkilä, Janne},
booktitle = {Proceedings - International Conference on Image Processing, ICIP}
}
Downloads: 0
{"_id":"KftQXTifEQucpd8dC","bibbaseid":"bayramoglu-kannala-akerfelt-kaakinen-eklund-nees-heikkil-anovelfeaturedescriptorbasedonmicroscopyimagestatistics-2015","authorIDs":[],"author_short":["Bayramoglu, N.","Kannala, J.","Akerfelt, M.","Kaakinen, M.","Eklund, L.","Nees, M.","Heikkilä, J."],"bibdata":{"title":"A novel feature descriptor based on microscopy image statistics","type":"inProceedings","year":"2015","identifiers":"[object Object]","keywords":"cell co-culture,cell detection,electron microscopy,local image descriptor,mitochondria,phase contrast imaging,pixel labeling,tumor","pages":"2695-2699","publisher":"IEEE","id":"06f66f46-06e4-3fb5-b62e-afaa9ebfaffe","created":"2019-09-15T16:34:27.675Z","file_attached":false,"profile_id":"bddcf02d-403b-3b06-9def-6d15cc293e20","last_modified":"2019-09-19T17:47:33.125Z","read":false,"starred":false,"authored":"true","confirmed":"true","hidden":false,"source_type":"CONF","private_publication":false,"abstract":"© 2015 IEEE. In this paper, we propose a novel feature description algorithm based on image statistics. The pipeline first performs independent component analysis on training image patches to obtain basis vectors (filters) for a lower dimensional representation. Then for a given image, a set of filter responses at each pixel is computed. Finally, a histogram representation, which considers the signs and magnitudes of the responses as well as the number of filters, is applied on local image patches. We propose to apply this idea to a microscopy image pixel identification system based on a learning framework. Experimental results show that the proposed algorithm performs better than the state-of-the-art descriptors in biomedical images of different microscopy modalities.","bibtype":"inProceedings","author":"Bayramoglu, Neslihan and Kannala, Juho and Akerfelt, Malin and Kaakinen, Mika and Eklund, Lauri and Nees, Matthias and Heikkilä, Janne","booktitle":"Proceedings - International Conference on Image Processing, ICIP","bibtex":"@inProceedings{\n title = {A novel feature descriptor based on microscopy image statistics},\n type = {inProceedings},\n year = {2015},\n identifiers = {[object Object]},\n keywords = {cell co-culture,cell detection,electron microscopy,local image descriptor,mitochondria,phase contrast imaging,pixel labeling,tumor},\n pages = {2695-2699},\n publisher = {IEEE},\n id = {06f66f46-06e4-3fb5-b62e-afaa9ebfaffe},\n created = {2019-09-15T16:34:27.675Z},\n file_attached = {false},\n profile_id = {bddcf02d-403b-3b06-9def-6d15cc293e20},\n last_modified = {2019-09-19T17:47:33.125Z},\n read = {false},\n starred = {false},\n authored = {true},\n confirmed = {true},\n hidden = {false},\n source_type = {CONF},\n private_publication = {false},\n abstract = {© 2015 IEEE. In this paper, we propose a novel feature description algorithm based on image statistics. The pipeline first performs independent component analysis on training image patches to obtain basis vectors (filters) for a lower dimensional representation. Then for a given image, a set of filter responses at each pixel is computed. Finally, a histogram representation, which considers the signs and magnitudes of the responses as well as the number of filters, is applied on local image patches. We propose to apply this idea to a microscopy image pixel identification system based on a learning framework. Experimental results show that the proposed algorithm performs better than the state-of-the-art descriptors in biomedical images of different microscopy modalities.},\n bibtype = {inProceedings},\n author = {Bayramoglu, Neslihan and Kannala, Juho and Akerfelt, Malin and Kaakinen, Mika and Eklund, Lauri and Nees, Matthias and Heikkilä, Janne},\n booktitle = {Proceedings - International Conference on Image Processing, ICIP}\n}","author_short":["Bayramoglu, N.","Kannala, J.","Akerfelt, M.","Kaakinen, M.","Eklund, L.","Nees, M.","Heikkilä, J."],"bibbaseid":"bayramoglu-kannala-akerfelt-kaakinen-eklund-nees-heikkil-anovelfeaturedescriptorbasedonmicroscopyimagestatistics-2015","role":"author","urls":{},"keyword":["cell co-culture","cell detection","electron microscopy","local image descriptor","mitochondria","phase contrast imaging","pixel labeling","tumor"],"downloads":0},"bibtype":"inProceedings","creationDate":"2019-09-19T17:48:29.221Z","downloads":0,"keywords":["cell co-culture","cell detection","electron microscopy","local image descriptor","mitochondria","phase contrast imaging","pixel labeling","tumor"],"search_terms":["novel","feature","descriptor","based","microscopy","image","statistics","bayramoglu","kannala","akerfelt","kaakinen","eklund","nees","heikkilä"],"title":"A novel feature descriptor based on microscopy image statistics","year":2015}