24小时热门版块排行榜    

查看: 989  |  回复: 5
当前只显示满足指定条件的回帖,点击这里查看本话题的所有回帖

郁闷的GENIUS

金虫 (小有名气)

[求助] 大家讨论下如何通过一组图像序列求取相机的移动轨迹啊? 已有1人参与

RT,小弟看了些论文,但过程讲的都不是很清楚,想请大神帮忙解答下:当两张图像的匹配点对求出来后,如何计算拍摄这两张图像时相机的相对位姿呢?
回复此楼
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

lxc_422

新虫 (小有名气)

【答案】应助回帖

Yes, you are right!
If you used Zhang's method to obtain the interal parameters of the camera(i.e. the calibration matrix K), then it is simple to solve this problem.
全心用于科研,不能有丝毫松懈。学术是主题,创新是生命。
4楼2014-12-25 12:50:37
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖
查看全部 6 个回答

lxc_422

新虫 (小有名气)

【答案】应助回帖

★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★
感谢参与,应助指数 +1
郁闷的GENIUS: 金币+5, ★★★很有帮助 2014-12-25 11:42:00
郁闷的GENIUS: 金币+10 2014-12-25 18:12:01
郁闷的GENIUS: 金币+10 2015-01-03 17:33:40
First, you should compute the essential matrix E.
Second, according to E=[t]xR, if you know the translation matrix t and the rotation matrix R, then you could obtain the pose of the camera.
However, you only have two images at hand. We assume that the camera are uncalibrated, therefore, you need to solve the fundamental matrix F from the corresponding points.
Since E=KT*F*K, you should compute the calibration matrix K. We know that K only could be determined from autocalibration(or self-calibration), because the camera has not been calibrated.
Hence, if you solved the problem of camera autocalibration and obtained K, then you could easily obtain the pose of the camera.
How could we solve the autocalibration problem? The idea of auto calibrating a camera originated in the literature titled "Camera self-calibration: Theory and experiments" by Faugeras, Luong and Maybank in 1992.
You could find out more useful information from this idea.
全心用于科研,不能有丝毫松懈。学术是主题,创新是生命。
2楼2014-12-25 10:09:32
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

郁闷的GENIUS

金虫 (小有名气)

引用回帖:
2楼: Originally posted by lxc_422 at 2014-12-25 10:09:32
First, you should compute the essential matrix E.
Second, according to E=xR, if you know the translation matrix t and the rotation matrix R, then you could obtain the pose of the camera.
However,  ...

非常感谢,如果先用张正友的方法将相机标定好,并得到了相关参数,是不是只要通过两张图像的匹配点对求本质矩阵E就可以了?
3楼2014-12-25 11:41:48
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

lxc_422

新虫 (小有名气)

【答案】应助回帖

I suggest you to add your question into the BBS.
It make others can read it, maybe this could help them while encountering the similar problem.
全心用于科研,不能有丝毫松懈。学术是主题,创新是生命。
5楼2014-12-25 12:55:57
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖
最具人气热帖推荐 [查看全部] 作者 回/看 最后发表
[基金申请] 面上合作单位盖章 +6 ssyjh 2026-08-27 9/450 2026-09-01 20:01 by huagongfeihu
[基金申请] 为什么国自然不能直接公布 +5 bjdxyxy 2026-08-26 5/250 2026-09-01 17:54 by 小伟大博士
[文学芳草园] 梦想 +5 myrtle 2026-08-26 7/350 2026-09-01 15:18 by myrtle
[基金申请] 学科评审组评审是指会评吗? +4 瞬息宇宙 2026-08-31 4/200 2026-09-01 14:58 by jiaoxg
[基金申请] 麻烦专家们看看评委们的意见(F口面上) +7 gdd2018 2026-08-28 12/600 2026-09-01 08:32 by 尼古拉斯小虫
[基金申请] 怎么看青基中了没有啊 +6 叶九微 2026-08-26 6/300 2026-08-31 23:54 by yudaoqian88
[基金申请] 国社科又开始会评了,不知道这次命运如何 +7 雨打竹帘 2026-08-30 11/550 2026-08-31 23:16 by hittle2008
[基金申请] 投票:  有多少人是今天查系统知道结果的? +17 爱看书的可乐 2026-08-26 19/950 2026-08-31 21:30 by xiangy672
[基金申请] 哪位高人中了,把查询到的截图贴出来让我看看,让我长长见识 +6 yuleib84 2026-08-26 7/350 2026-08-31 19:46 by 鱼翔浅底1
[基金申请] 面上意见出来了 +12 黄鸟于飞Chao 2026-08-29 23/1150 2026-08-31 18:57 by 黄鸟于飞Chao
[基金申请] 29号明天会评吗 +4 笨笨唐 2026-08-28 4/200 2026-08-31 09:30 by huixian257
[基金申请] 基金不中,共勉 +12 eulota 2026-08-26 12/600 2026-08-31 08:42 by ZJTJZ
[基金申请] 为什么到现在没收到通知? +5 tannykie 2026-08-29 5/250 2026-08-30 21:05 by purplejack
[基金申请] 有没有仍没收到信息的 +7 德尚中行 2026-08-27 8/400 2026-08-30 20:52 by purplejack
[考博] 找导师 +6 yuanjiabao 2026-08-29 7/350 2026-08-30 14:40 by 生科新手
[基金申请] 2026年叶企孙基金 +4 bud_bud 2026-08-27 7/350 2026-08-29 07:23 by foolishmani
[基金申请] 为什么 国际(地区)合作与交流项目 没有放榜? 10+3 majunge000 2026-08-26 11/550 2026-08-27 08:42 by 北京莱茵编辑
[基金申请] 出来了 +9 trojank 2026-08-26 9/450 2026-08-26 14:25 by 宝贝虫子
[基金申请] 国合里面能看到了 +7 一怀馨秋 2026-08-26 7/350 2026-08-26 11:23 by zhaosm1982
[基金申请] 牛来!米来!面来! +8 beefly 2026-08-26 8/400 2026-08-26 08:37 by xuzhipiao
信息提示
请填处理意见