slam复习准备大纲

视觉前端

SLAM复习总结：视觉里程计前端总结(一)

• 视差几何模型推导(P91)；
  
  $$\begin{gathered} \frac{z-f}{z}=\frac{b-u_L+u_R}{b} \\ z=\frac{fb}{d},d=u_L-u_R \end{gathered}$$
• slam问题建模公式(P106)；
  $$\{\begin{array}{l}{\mathbf{x}}_k=f\left({\mathbf{x}}_{k-1},{\mathbf{u}}_k\right)+{\mathbf{w}}_k\\ {\mathbf{z}}_k=h\left({\mathbf{x}}_k\right)+{\mathbf{v}}_k\end{array}k=1,\dots ,N$$
• 第四讲公式：
  $$\begin{gathered} \exp \left(\Delta {\mathbf{\varphi }}^{\wedge }\right)\exp \left({\mathbf{\varphi }}^{\wedge }\right)=\exp \left({\left(\mathbf{\varphi }+{\mathbf{J}}_l^{-1}(\mathbf{\varphi })\Delta \mathbf{\varphi }\right)}^{\wedge }\right) \\ \exp \left((\mathbf{\varphi }+\Delta \mathbf{\varphi }{)}^{\wedge }\right)=\exp \left({\left({\mathbf{J}}_l\Delta \mathbf{\varphi }\right)}^{\wedge }\right)\exp \left({\mathbf{\varphi }}^{\wedge }\right)=\exp \left({\mathbf{\varphi }}^{\wedge }\right)\exp \left({\left({\mathbf{J}}_r\Delta \mathbf{\varphi }\right)}^{\wedge }\right) \end{gathered}$$
  李代数求导
  $$\begin{array}{rl}\frac{\partial (\mathbf{R}\mathbf{p})}{\partial \mathbf{\phi }}& =\underset{\phi \to 0}{\lim }\frac{\exp \left({\mathbf{\phi }}^{\wedge }\right)\exp \left({\mathbf{\varphi }}^{\wedge }\right)\mathbf{p}-\exp \left({\mathbf{\varphi }}^{\wedge }\right)\mathbf{p}}{\mathbf{\phi }}\\ & \approx \underset{\phi \to 0}{\lim }\frac{\left(1+{\mathbf{\phi }}^{\wedge }\right)\exp \left({\mathbf{\varphi }}^{\wedge }\right)\mathbf{p}-\exp \left({\mathbf{\varphi }}^{\wedge }\right)\mathbf{p}}{\mathbf{\phi }}\\ & =\underset{\phi \to 0}{\lim }\frac{{\mathbf{\phi }}^{\wedge }\mathbf{R}\mathbf{p}}{\mathbf{\phi }}=\underset{\phi \to 0}{\lim }\frac{-(\mathbf{R}\mathbf{p}{)}^{\wedge }\mathbf{\phi }}{\mathbf{\phi }}=-(\mathbf{R}\mathbf{p}{)}^{\wedge }\end{array}$$
• SIFT算法详解
• LOG高斯-拉普拉斯算子
• 尺度不变特征变换(SIFT)特征提取分析
• SURF特征提取分析
• SIFT算法
• Programming Computer Vision with Python （学习笔记十一）
• 学习笔记 2.1 — Harris角点检测与特征匹配【含实例】
• 光流
  $$\begin{gathered} \mathbf{I}(x+\mathrm{d}x,y+\mathrm{d}y,t+\mathrm{d}t)\approx \mathbf{I}(x,y,t)+\frac{\partial \mathbf{I}}{\partial x}\mathrm{d}x+\frac{\partial \mathbf{I}}{\partial y}\mathrm{d}y+\frac{\partial \mathbf{I}}{\partial t}\mathrm{d}t \\ \frac{\partial \mathbf{I}}{\partial x}\mathrm{d}x+\frac{\partial \mathbf{I}}{\partial y}\mathrm{d}y+\frac{\partial \mathbf{I}}{\partial t}\mathrm{d}t=0 \\ \frac{\partial \mathbf{I}}{\partial x}\frac{\mathrm{d}x}{\mathrm{d}t}+\frac{\partial \mathbf{I}}{\partial y}\frac{\mathrm{d}y}{\mathrm{d}t}=-\frac{\partial \mathbf{I}}{\partial t} \\ \left[\begin{array}{ll}{\mathbf{I}}_x& {\mathbf{I}}_y\end{array}\right]\left[\begin{array}{l}u\\ v\end{array}\right]=-{\mathbf{I}}_t \\ \mathbf{A}\left[\begin{array}{l}u\\ v\end{array}\right]=-\mathbf{b} \\ {\left[\begin{array}{l}u\\ v\end{array}\right]}^{\ast }=-{\left({\mathbf{A}}^T\mathbf{A}\right)}^{-1}{\mathbf{A}}^T\mathbf{b} \end{gathered}$$

后端优化

P ( x k ∣ x 0 , u 1 : k , z 1 : k ) P ( x k ∣ x 0 , u 1 : k , z 1 : k ) ∝ P ( z k ∣ x k ) P ( x k ∣ x 0 , u 1 : k , z 1 : k − 1 ) P ( x k ∣ x 0 , u 1 : k , z 1 : k − 1 ) = ∫ P ( x k ∣ x k − 1 , x 0 , u 1 : k , z 1 : k − 1 ) P ( x k − 1 ∣ x 0 , u 1 : k , z 1 : k − 1 ) d x k − 1 P ( x k ∣ x k − 1 , x 0 , u 1 : k , z 1 : k − 1 ) = P ( x k ∣ x k − 1 , u k ) P ( x k − 1 ∣ x 0 , u 1 : k , z 1 : k − 1 ) = P ( x k − 1 ∣ x 0 , u 1 : k − 1 , z 1 : k − 1 ) { x k = A k x k − 1 + u k + w k z k = C k x k + v k k = 1 , … , N w k ∼ N ( 0 , R ) . v k ∼ N ( 0 , Q ) P\left(\boldsymbol{x}_{k} | \boldsymbol{x}_{0}, \boldsymbol{u}_{1 : k}, \boldsymbol{z}_{1 : k}\right) \\P\left(\boldsymbol{x}_{k} | \boldsymbol{x}_{0}, \boldsymbol{u}_{1 : k}, \boldsymbol{z}_{1 : k}\right) \propto P\left(\boldsymbol{z}_{k} | \boldsymbol{x}_{k}\right) P\left(\boldsymbol{x}_{k} | \boldsymbol{x}_{0}, \boldsymbol{u}_{1 : k}, \boldsymbol{z}_{1 : k-1}\right)\\P\left(\boldsymbol{x}_{k} | \boldsymbol{x}_{0}, \boldsymbol{u}_{1 : k}, \boldsymbol{z}_{1 : k-1}\right)=\int P\left(\boldsymbol{x}_{k} | \boldsymbol{x}_{k-1}, \boldsymbol{x}_{0}, \boldsymbol{u}_{1 : k}, \boldsymbol{z}_{1 : k-1}\right) P\left(\boldsymbol{x}_{k-1} | \boldsymbol{x}_{0}, \boldsymbol{u}_{1 : k}, \boldsymbol{z}_{1 : k-1}\right) \mathrm{d} \boldsymbol{x}_{k-1}\\P\left(\boldsymbol{x}_{k} | \boldsymbol{x}_{k-1}, \boldsymbol{x}_{0}, \boldsymbol{u}_{1 : k}, \boldsymbol{z}_{1 : k-1}\right)=P\left(\boldsymbol{x}_{k} | \boldsymbol{x}_{k-1}, \boldsymbol{u}_{k}\right)\\P\left(\boldsymbol{x}_{k-1} | \boldsymbol{x}_{0}, \boldsymbol{u}_{1 : k}, \boldsymbol{z}_{1 : k-1}\right)=P\left(\boldsymbol{x}_{k-1} | \boldsymbol{x}_{0}, \boldsymbol{u}_{1 : k-1}, \boldsymbol{z}_{1 : k-1}\right)\\\left\{\begin{array}{ll}{\boldsymbol{x}_{k}=\boldsymbol{A}_{k} \boldsymbol{x}_{k-1}+\boldsymbol{u}_{k}+\boldsymbol{w}_{k}} \\ {\boldsymbol{z}_{k}=\boldsymbol{C}_{k} \boldsymbol{x}_{k}+\boldsymbol{v}_{k}}\end{array} \quad k=1, \ldots, N\right.\\\boldsymbol{w}_{k} \sim N(\mathbf{0}, \boldsymbol{R}) . \quad \boldsymbol{v}_{k} \sim N(\mathbf{0}, \boldsymbol{Q})