{ "id": 488325, "title": "Question regarding Rotation matrix and Translation vector", "url": "/competitions/image-matching-challenge-2024/discussion/488325", "author_name": "", "post_date": "2024-04-02T03:26:21.162294300Z", "votes": 4, "comment_count": 4, "views": 0, "content": "

Can anyone please tell me what does the rotation matrix(R) and translation vector(t) in the training set represent. Does it tell about the transformation from one camera position to another or from 3D object to camera? Where is the (0, 0, 0) coordinate located?

", "messages": [ { "id": "2727987", "postDate": "04/02/2024 03:26:21", "content": "

Can anyone please tell me what does the rotation matrix(R) and translation vector(t) in the training set represent. Does it tell about the transformation from one camera position to another or from 3D object to camera? Where is the (0, 0, 0) coordinate located?

", "rawMarkdown": "Can anyone please tell me what does the rotation matrix(R) and translation vector(t) in the training set represent. Does it tell about the transformation from one camera position to another or from 3D object to camera? Where is the (0, 0, 0) coordinate located?", "votes": null }, { "id": "2728371", "postDate": "04/02/2024 07:39:41", "content": "

Hi, R and t follows the conventions of COLMAP, you can have a look here https://colmap.github.io/format.html#images-txt

", "rawMarkdown": "Hi, R and t follows the conventions of COLMAP, you can have a look here https://colmap.github.io/format.html#images-txt", "votes": null }, { "id": "2728830", "postDate": "04/02/2024 12:27:30", "content": "

Thank you for your reply. According to the link the Rt transformation represents the projection from world to camera coordinate.
\nThe reconstructed pose of an image is specified as the projection from world to the camera coordinate system of an image using a quaternion (QW, QX, QY, QZ) and a translation vector (TX, TY, TZ).

\n

But I am still confused where the world coordinate system is. Every image sees different parts of the object.

", "rawMarkdown": "Thank you for your reply. According to the link the Rt transformation represents the projection from world to camera coordinate.\n`The reconstructed pose of an image is specified as the projection from world to the camera coordinate system of an image using a quaternion (QW, QX, QY, QZ) and a translation vector (TX, TY, TZ).`\n\nBut I am still confused where the world coordinate system is. Every image sees different parts of the object.", "votes": null }, { "id": "2729326", "postDate": "04/02/2024 16:37:10", "content": "

T is the vector that from the ref system of each camera points to the world reference system. To have the position of each camera in the world reference system you should do -R^t * T

", "rawMarkdown": "T is the vector that from the ref system of each camera points to the world reference system. To have the position of each camera in the world reference system you should do `-R^t * T`", "votes": null }, { "id": "2732161", "postDate": "04/03/2024 02:58:55", "content": "

now, I understand. Thank you so much for your detailed explanation.

", "rawMarkdown": "now, I understand. Thank you so much for your detailed explanation.", "votes": null } ], "comments": [ { "id": 2728371, "author_name": "lcmrll", "author_url": "", "post_date": "04/02/2024 07:39:41", "content": "

Hi, R and t follows the conventions of COLMAP, you can have a look here https://colmap.github.io/format.html#images-txt

", "votes": null, "replies": [ { "id": 2728830, "author_name": "tabassumnova", "author_url": "", "post_date": "04/02/2024 12:27:30", "content": "

Thank you for your reply. According to the link the Rt transformation represents the projection from world to camera coordinate.
\nThe reconstructed pose of an image is specified as the projection from world to the camera coordinate system of an image using a quaternion (QW, QX, QY, QZ) and a translation vector (TX, TY, TZ).

\n

But I am still confused where the world coordinate system is. Every image sees different parts of the object.

", "votes": null, "replies": [ { "id": 2729326, "author_name": "lcmrll", "author_url": "", "post_date": "04/02/2024 16:37:10", "content": "

T is the vector that from the ref system of each camera points to the world reference system. To have the position of each camera in the world reference system you should do -R^t * T

", "votes": null, "replies": [ { "id": 2732161, "author_name": "tabassumnova", "author_url": "", "post_date": "04/03/2024 02:58:55", "content": "

now, I understand. Thank you so much for your detailed explanation.

", "votes": null, "replies": [] } ] } ] } ] } ], "raw_markdown_by_id": { "2727987": "Can anyone please tell me what does the rotation matrix(R) and translation vector(t) in the training set represent. Does it tell about the transformation from one camera position to another or from 3D object to camera? Where is the (0, 0, 0) coordinate located?", "2728371": "Hi, R and t follows the conventions of COLMAP, you can have a look here https://colmap.github.io/format.html#images-txt", "2728830": "Thank you for your reply. According to the link the Rt transformation represents the projection from world to camera coordinate.\n`The reconstructed pose of an image is specified as the projection from world to the camera coordinate system of an image using a quaternion (QW, QX, QY, QZ) and a translation vector (TX, TY, TZ).`\n\nBut I am still confused where the world coordinate system is. Every image sees different parts of the object.", "2729326": "T is the vector that from the ref system of each camera points to the world reference system. To have the position of each camera in the world reference system you should do `-R^t * T`", "2732161": "now, I understand. Thank you so much for your detailed explanation." }, "source": "meta" }