Coherent Motion Segmentation in Moving Camera Videos using Optical Flow Orientations. Narayana, M. & Hanson, A. ![Coherent Motion Segmentation in Moving Camera Videos using Optical Flow Orientations [pdf]](https://bibbase.org/img/filetypes/pdf.svg "pdf")
Paper abstract bibtex In moving camera videos, motion segmentation is com-monly performed using the image plane motion of pixels, or optical flow. However, objects that are at different depths from the camera can exhibit different optical flows even if they share the same real-world motion. This can cause a depth-dependent segmentation of the scene. Our goal is to develop a segmentation algorithm that clusters pixels that have similar real-world motion irrespective of their depth in the scene. Our solution uses optical flow orientations in-stead of the complete vectors and exploits the well-known property that under camera translation, optical flow ori-entations are independent of object depth. We introduce a probabilistic model that automatically estimates the number of observed independent motions and results in a labeling that is consistent with real-world motion in the scene. The result of our system is that static objects are correctly iden-tified as one segment, even if they are at different depths. Color features and information from previous frames in the video sequence are used to correct occasional errors due to the orientation-based segmentation. We present results on more than thirty videos from different benchmarks. The system is particularly robust on complex background scenes containing objects at significantly different depths.
@article{
title = {Coherent Motion Segmentation in Moving Camera Videos using Optical Flow Orientations},
type = {article},
identifiers = {[object Object]},
id = {46fa0a0d-aa92-3ca3-98b9-b924fe9d5c32},
created = {2017-08-16T18:53:45.349Z},
file_attached = {true},
profile_id = {47354232-dc6c-341e-8f5d-b4fb19349495},
group_id = {2b397b4f-2cb3-330b-a507-8e6163be955c},
last_modified = {2017-08-16T18:53:46.773Z},
read = {false},
starred = {false},
authored = {false},
confirmed = {false},
hidden = {false},
folder_uuids = {fc79bdea-30ac-481c-8bf0-fae6a1a5be1a},
private_publication = {false},
abstract = {In moving camera videos, motion segmentation is com-monly performed using the image plane motion of pixels, or optical flow. However, objects that are at different depths from the camera can exhibit different optical flows even if they share the same real-world motion. This can cause a depth-dependent segmentation of the scene. Our goal is to develop a segmentation algorithm that clusters pixels that have similar real-world motion irrespective of their depth in the scene. Our solution uses optical flow orientations in-stead of the complete vectors and exploits the well-known property that under camera translation, optical flow ori-entations are independent of object depth. We introduce a probabilistic model that automatically estimates the number of observed independent motions and results in a labeling that is consistent with real-world motion in the scene. The result of our system is that static objects are correctly iden-tified as one segment, even if they are at different depths. Color features and information from previous frames in the video sequence are used to correct occasional errors due to the orientation-based segmentation. We present results on more than thirty videos from different benchmarks. The system is particularly robust on complex background scenes containing objects at significantly different depths.},
bibtype = {article},
author = {Narayana, Manjunath and Hanson, Allen}
}
Downloads: 0
{"_id":"vQCoYN94ohFKYRwbf","bibbaseid":"narayana-hanson-coherentmotionsegmentationinmovingcameravideosusingopticalfloworientations","downloads":0,"creationDate":"2017-10-26T16:18:17.780Z","title":"Coherent Motion Segmentation in Moving Camera Videos using Optical Flow Orientations","author_short":["Narayana, M.","Hanson, A."],"year":null,"bibtype":"article","biburl":null,"bibdata":{"title":"Coherent Motion Segmentation in Moving Camera Videos using Optical Flow Orientations","type":"article","identifiers":"[object Object]","id":"46fa0a0d-aa92-3ca3-98b9-b924fe9d5c32","created":"2017-08-16T18:53:45.349Z","file_attached":"true","profile_id":"47354232-dc6c-341e-8f5d-b4fb19349495","group_id":"2b397b4f-2cb3-330b-a507-8e6163be955c","last_modified":"2017-08-16T18:53:46.773Z","read":false,"starred":false,"authored":false,"confirmed":false,"hidden":false,"folder_uuids":"fc79bdea-30ac-481c-8bf0-fae6a1a5be1a","private_publication":false,"abstract":"In moving camera videos, motion segmentation is com-monly performed using the image plane motion of pixels, or optical flow. However, objects that are at different depths from the camera can exhibit different optical flows even if they share the same real-world motion. This can cause a depth-dependent segmentation of the scene. Our goal is to develop a segmentation algorithm that clusters pixels that have similar real-world motion irrespective of their depth in the scene. Our solution uses optical flow orientations in-stead of the complete vectors and exploits the well-known property that under camera translation, optical flow ori-entations are independent of object depth. We introduce a probabilistic model that automatically estimates the number of observed independent motions and results in a labeling that is consistent with real-world motion in the scene. The result of our system is that static objects are correctly iden-tified as one segment, even if they are at different depths. Color features and information from previous frames in the video sequence are used to correct occasional errors due to the orientation-based segmentation. We present results on more than thirty videos from different benchmarks. The system is particularly robust on complex background scenes containing objects at significantly different depths.","bibtype":"article","author":"Narayana, Manjunath and Hanson, Allen","bibtex":"@article{\n title = {Coherent Motion Segmentation in Moving Camera Videos using Optical Flow Orientations},\n type = {article},\n identifiers = {[object Object]},\n id = {46fa0a0d-aa92-3ca3-98b9-b924fe9d5c32},\n created = {2017-08-16T18:53:45.349Z},\n file_attached = {true},\n profile_id = {47354232-dc6c-341e-8f5d-b4fb19349495},\n group_id = {2b397b4f-2cb3-330b-a507-8e6163be955c},\n last_modified = {2017-08-16T18:53:46.773Z},\n read = {false},\n starred = {false},\n authored = {false},\n confirmed = {false},\n hidden = {false},\n folder_uuids = {fc79bdea-30ac-481c-8bf0-fae6a1a5be1a},\n private_publication = {false},\n abstract = {In moving camera videos, motion segmentation is com-monly performed using the image plane motion of pixels, or optical flow. However, objects that are at different depths from the camera can exhibit different optical flows even if they share the same real-world motion. This can cause a depth-dependent segmentation of the scene. Our goal is to develop a segmentation algorithm that clusters pixels that have similar real-world motion irrespective of their depth in the scene. Our solution uses optical flow orientations in-stead of the complete vectors and exploits the well-known property that under camera translation, optical flow ori-entations are independent of object depth. We introduce a probabilistic model that automatically estimates the number of observed independent motions and results in a labeling that is consistent with real-world motion in the scene. The result of our system is that static objects are correctly iden-tified as one segment, even if they are at different depths. Color features and information from previous frames in the video sequence are used to correct occasional errors due to the orientation-based segmentation. We present results on more than thirty videos from different benchmarks. The system is particularly robust on complex background scenes containing objects at significantly different depths.},\n bibtype = {article},\n author = {Narayana, Manjunath and Hanson, Allen}\n}","author_short":["Narayana, M.","Hanson, A."],"urls":{"Paper":"http://bibbase.org/service/mendeley/4433140d-2382-3a4a-a3e6-b86f672ad203/file/7063bee0-52f7-65d9-3ce0-4c1a3795f34e/Coherent_Motion_Segmentation_in_Moving_Camera_Videos_using_Optical_Flow_Orientations.pdf.pdf"},"bibbaseid":"narayana-hanson-coherentmotionsegmentationinmovingcameravideosusingopticalfloworientations","role":"author","downloads":0,"html":""},"search_terms":["coherent","motion","segmentation","moving","camera","videos","using","optical","flow","orientations","narayana","hanson"],"keywords":[],"authorIDs":[]}