Paper Reading: Stereo DSO
开篇第一篇就写一个paper reading吧,用markdown+vim写东西切换中英文挺麻烦的,有些就偷懒都用英文写了。
Stereo DSO: Large-Scale Direct Sparse Visual Odometry with Stereo Cameras
Abstract
Optimization objectives:
- intrinsic/extrinsic parameters of all keyframes
- all selected pixels' depth
Integrate constraints from static stereo (左右两个相机的立体视觉约束是静态的) into the bundle adjustment pipeline of temporal multi-view stereo.
Fixed-baseline stereo resolves scale drift.
? It also reduces the sensitivities to large optical flow and to rolling shutter effect which are known shortcomings of direct image alignment methods.
1. Introduction
stem from: working in an effective way
heuristically: 启发式的
hallucinate: 出现幻觉
strip down: reduced to its simplest form
Strasdat et al. proposed to expand the concept of keyframes to integrate scale and proposed a double window optimization (Figure out what is it)
Direct methods aim at computing geometry and motion directly from the images thereby skipping the intermediate keypoint selection step.
The key idea of LSD SLAM is to incrementally track the camera and simultaneously perform a pose graph optimization in order to keep the entire camera trajectory globally consistent. 作者认为这种方式没有减少累计误差,只是把它扩散到整个轨迹中( So the meaning of pose graph is? )。
Three drawbacks of DSO:
- The mentioned performance was gained on a photometrically calibrated dataset, in its absense, the performance would degrade.
- Scale drift
- DSO is quite sensitive to geometric distortion as those induces by fast motion and rolling shutter. While techniques for calibrating rolling shutter exist for direct SLAM algorithm, these are often quite involved and far from real-time capable.
Contribution:
- A stereo version of DSO. detail the proposed combination of temporal multi-view stereo and static stereo.
- Stereo DSO is good.
2. Direct Sparse VO with Stereo Camera
- Absolute scale can be directly calculated from static stereo from the known baseline of the stereo camera
- Static stereo can provide initial depth estimation for multi-view stereo
- Static Stereo can only accurately triangulate 3D points within a limited depth range while this limit is resolved by temporal multi-view stereo.
New stereo frames are first tracked with respect to their reference keyframe in a coarse-to-fine mannar.
A joint optimization of their poses, affine brightness (两个参数:a和b) parameters, as well as the depts of all the observed 3D points and camera intrinsics, is performed.
2.1 Notation
Nothing important.
2.2 Direct Image Alignment Formulation
\[
E_{ij}=\sum_{p\in P_i}\omega_p \left\| I_j[p'] - I_i[p] \right\|_\gamma
\]
where \(\omega_p\) is the weight which is shown as follows.(梯度越大权重越小,不知道为啥)
\[
\omega_p = \frac{c^2}{c^2+\left\| \nabla I_i(p) \right\| ^2_2}
\]
光度误差对突然的光照变化非常敏感。
2.3 Tracking
All the potins inside the active window are projected into the new frame. Then the pose of the new frame is optimized by minimizing the energy function.
在之前的单目DSO中,用随机深度值来初始化,所以都会需要一个确定模式的移动来初始化。在本文中,因为这时候stereo image pair的affine brightness transfer factor是位置的,所以用NCC在水平极限上的3*5的领域中搜索。
2.4 Frame Management
The basic idea is to check if the scene or the illumination has sufficiently changed.
- scene change: 用mean square optical flow和 mean squared optical flow without rotation between the current frame and the last keyframe来衡量。
- illumination change: 用relative brightness factor \(|a_j - a_i|\) 来衡量。
-> 一个点如果是梯度大于一个阈值并且是一个block里最大的点,那么他会被选择。
-> Before a candidate point is activated and optimized in the windowed optimization, its inverse depth is constantly refined by the following non-keyframes. (找出来怎么做的)
-> 旧去新来:在边缘化点的时候把候选点加入到联合优化中。
-> The constraints from static stereo introduce scale information into the system, and they also provide good geometric priors to temporal multi-view stereo.
2.5 Windowed Optimization
-> Temporal Multi-View Stereo: 就一般的不同时刻的图片之间的立体视觉
-> Static Stereo:
-> Stereo Coupling: 为了平衡上两种约束的权重,我们引入了\(\lambda\)参数。
-> Margninalization: 在边缘化一个关键帧之前,我们首先会边缘化所有没有被过去两个关键帧看到所有active window中的点。
3. Evaluation
暂且略过不表
4. Conclusion
未来可以做的两件事:
- Loop closuring and a database for map maintenance (LDSO半闲居士做过了)
- Dynamic object handling to further boost the VO accuracy and robustness. (用深度学习做动态物体检测然后动的点不要了?)
虽然自己在SLAM领域还有很多可以学习的,但是这样感觉直接法的东西也做完了?悲伤。。
Paper Reading: Stereo DSO的更多相关文章
- [Paper Reading]--Exploiting Relevance Feedback in Knowledge Graph
<Exploiting Relevance Feedback in Knowledge Graph> Publication: KDD 2015 Authors: Yu Su, Sheng ...
- Paper Reading: Perceptual Generative Adversarial Networks for Small Object Detection
Perceptual Generative Adversarial Networks for Small Object Detection 2017-07-11 19:47:46 CVPR 20 ...
- Paper Reading: In Defense of the Triplet Loss for Person Re-Identification
In Defense of the Triplet Loss for Person Re-Identification 2017-07-02 14:04:20 This blog comes ...
- Paper Reading - Attention Is All You Need ( NIPS 2017 ) ★
Link of the Paper: https://arxiv.org/abs/1706.03762 Motivation: The inherently sequential nature of ...
- Paper Reading - Convolutional Sequence to Sequence Learning ( CoRR 2017 ) ★
Link of the Paper: https://arxiv.org/abs/1705.03122 Motivation: Compared to recurrent layers, convol ...
- Paper Reading - Deep Captioning with Multimodal Recurrent Neural Networks ( m-RNN ) ( ICLR 2015 ) ★
Link of the Paper: https://arxiv.org/pdf/1412.6632.pdf Main Points: The authors propose a multimodal ...
- Paper Reading - Deep Visual-Semantic Alignments for Generating Image Descriptions ( CVPR 2015 )
Link of the Paper: https://arxiv.org/abs/1412.2306 Main Points: An Alignment Model: Convolutional Ne ...
- Paper Reading - Mind’s Eye: A Recurrent Visual Representation for Image Caption Generation ( CVPR 2015 )
Link of the Paper: https://ieeexplore.ieee.org/document/7298856/ A Correlative Paper: Learning a Rec ...
- Paper Reading - Show and Tell: A Neural Image Caption Generator ( CVPR 2015 )
Link of the Paper: https://arxiv.org/abs/1411.4555 Main Points: A generative model ( NIC, GoogLeNet ...
随机推荐
- TeamCity 创建jar构建步骤
1 创建工程 2 配置工程代码来源信息 2.1 From a repository URL 表示从代码仓库创建工程. 2.1.1 parent project 指定父工程,默认是root projec ...
- MTK-TP(触屏)解读一
MTK中的TP代码结构并不复杂,相比于其他的系统更为的简单些.它使用的是input子系统,通过该系统来上报触摸按键. 首先我们来看看TP的文件夹下的各代码文件的功能. 文件名 具体功能 关系文件 tp ...
- zeros()和ones()和eye()
python--zeros函数和ones函数 使用numpy.zeros,numpy.ones,numpy.eye等方法可以构造特定的矩阵 >>>from numpy import ...
- 查看mysql中sql语句执行时间
查看mysql版本:select version();方法一: show profiles.1. Show profiles是5.0.37之后添加的,要想使用此功能,要确保版本在5.0.37之后. ...
- eclipse改jsp文件编码格式 统一设置
- (转) NAS(神经结构搜索)综述
NAS(神经结构搜索)综述 文章转载自:http://www.tensorinfinity.com/paper_136.html 本文是对神经结构搜索(NAS)的简单综述,在写作的过程中参考了文献[1 ...
- SQL server中如何按照某一字段中的分割符将记录拆成多条
现需要将上结果转换为下结果 上结果查询语句:SELECT TOP 1 id,domain FROM dbo.SimpleTask 下结果转换语句:SELECT a.Id,b.domain FROM ...
- git\CentOS6.5中gitlab安装教程
一.Git 起源: Git是一个开源的分布式版本控制系统,用于敏捷高效地处理任何或小或大的项目. Git 是 Linus Torvalds 为了帮助管理 Linux 内核开发而开发的一个开放源码的版本 ...
- Python中常见的正则表达式符号
? 匹配零次或一次前面的分组 * 匹配零次或多次前面的分组 + 匹配一次或多次前面的分组 {n} 匹配n次前面的分组 {n,} 匹配n次或更多次前面的分组 {,m} 匹配零次到m次前面的分组 ...
- java @FunctionalInterface
public class Worker { private String doWork(String job) { return "Job " + job + " don ...