![Page 1: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/1.jpg)
CS231A Midterm Review
Friday 5/6/2016
![Page 2: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/2.jpg)
Outline• General Logistics • Camera Models • Non-perspective cameras • Calibration • Single View Metrology • Epipolar Geometry • Structure from Motion • Active Stereo and Volumetric Stereo • Fitting and Matching
• RANSAC • Hough Transform
• Detectors and Descriptors
![Page 3: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/3.jpg)
Midterm Logistics• In class midterm at Skilling Aud. 3:00pm-4:20pm on
Monday 5/9/2016
• SCPD students not taking exam at Stanford should have already set up a proctor.
• Exam covers material from lecture 1-10 (through detectors and descriptors)
• Approximately 10 TF, 5 MC, 8 Short Answer.
• Open book, open notes, open computer without network access
![Page 4: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/4.jpg)
Camera Models
![Page 5: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/5.jpg)
![Page 6: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/6.jpg)
![Page 7: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/7.jpg)
Transformations in 2D
Isometric
Similarity
Affine
Projective
![Page 8: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/8.jpg)
Properties of projective transformations • points project to points • lines project to lines • distant objects look smaller • angles are not preserved • parallel lines in 3D meet in the image
![Page 9: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/9.jpg)
Non-perspective cameras• Orthographic
• Weak Perspective
• Affine
![Page 10: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/10.jpg)
Calibration Problem
![Page 11: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/11.jpg)
Calibration Problem
![Page 12: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/12.jpg)
Calibration Problem
![Page 13: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/13.jpg)
Calibration Problem
![Page 14: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/14.jpg)
Homogeneous MxN Linear System
![Page 15: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/15.jpg)
Any real mxn matrix A can be decomposed uniquely as where U & V are orthogonal and is diagonal. If A has rank r,
Be familiar with how to use SVD to solve Ax=0 and Ax=b
Quick SVD recap
V = [v1, v2, ..., vr]U = [u1, u2, ..., ur]⌃r = diag(�1,�2, ...,�r)
A = U⌃V T ⌃
A = Ur⌃rVTr =
rX
i=1
�iuivTi
![Page 16: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/16.jpg)
Single View Metrology
• Vanishing points
• Vanishing lines
• Construction of lines from points
• Directions and normals of vanishing points and vanishing lines.
![Page 17: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/17.jpg)
Lines in a 2D plane
![Page 18: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/18.jpg)
Points at infinity
x1Point at infinity
Intersection of two lines
l1Line at infinity
Vanishing point:
the projective projection of a point at infinity into the image
plane
x = l ⇥ l
0
Line through two pointsl = x⇥ x
0
![Page 19: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/19.jpg)
Vanishing points and directions
![Page 20: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/20.jpg)
Vanishing lines
![Page 21: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/21.jpg)
• As we saw in problem set 1, we can calculate the internal camera matrix K provided a single view and knowledge of the the scene geometry.
Single View Calibration
![Page 22: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/22.jpg)
Multi-view Geometry• Camera Geometry: given corresponding points
in two images, find camera matrices, position and pose.
• Scene Geometry: Find correspondences of 3D points from its projection into 2 or more images.
• Correspondence: Given a point p in one image, how can I find the corresponding point p’ in another?
![Page 23: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/23.jpg)
Epipolar Geometry
![Page 24: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/24.jpg)
Epipolar Geometry
![Page 25: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/25.jpg)
Fundamental MatrixF = K�T
1 · [Tx
] ·RK�12
pT1 · Fp2 = 0
l1 = Fp2l2 = Fp1
F is 3x3 matrix; 7 DOF
F is singular (rank 2)
![Page 26: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/26.jpg)
Epipolar Geometry
Only need F to establish a relationship between the two corresponding points in the image No knowledge of position of P in 3D, or intrinsic/extrinsic parameters.
![Page 27: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/27.jpg)
8-point algorithm
![Page 28: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/28.jpg)
8-point algorithm
![Page 29: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/29.jpg)
Solving the Correspondenceissues: occlusion illumination foreshortening homogeneous regions repetitive patterns
![Page 30: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/30.jpg)
Structure From Motion
Structure = 3D pointsMotion = projection matrices
![Page 31: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/31.jpg)
Affine Structure From MotionAffine camera
xij = AiXj + bi
Problem: estimate the m matrices Ai, m matrices bi, and the n positions Xj from the mxn observations xij.
x̂ij = AiX̂j normalized
![Page 32: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/32.jpg)
Affine Structure From MotionAffine camera
xij = AiXj + bi
x̂ij = AiX̂j normalized
Factorization method
![Page 33: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/33.jpg)
Factorization methoduse SVD!
D should be rank 3
![Page 34: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/34.jpg)
Structure From Motion
Structure = 3D pointsMotion = projection matrices
![Page 35: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/35.jpg)
Structure From Motion
Structure = 3D pointsMotion = projection matrices
1. Recover structure and motion up to perspective ambiguity
• Algebraic approach using the Fundamental matrix
• Factorization method (by SVD) • Bundle adjustment
2. Resolving the perspective ambiguity
![Page 36: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/36.jpg)
Structure From MotionLimitations: 1. Algebraic approach using the Fundamental matrix
• Yields pairwise solutions. 2. Factorization method (by SVD)
• Assumes that all points are visible in every image, doesn’t handle occlusions.
Bundle adjustment via optimization • Attempts to minimize the re-projection error (pixel
distance between the projection of a reconstructed point into the estimated cameras for all the cameras and all the points).
![Page 37: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/37.jpg)
Active Stereo and Volumetric Stereo
Active Stereo • Structured lighting • Depth Sensing
Volumetric Stereo • Space carving • Shadow carving • Voxel coloring
![Page 38: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/38.jpg)
Active Stereo
![Page 39: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/39.jpg)
Volumetric Stereo
![Page 40: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/40.jpg)
Volumetric Stereo
![Page 41: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/41.jpg)
Volumetric StereoSpace Carving • remove voxels
via multiple views.
![Page 42: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/42.jpg)
Fitting and Matching• Least Squares methods • RANSAC • Hough transform
![Page 43: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/43.jpg)
FittingGoal: Find the best model parameters to fit the data.
Things we might want to fit include • Lines • Curves • Homographic transformations • Fundamental matrices
Issues with Fitting • Noisy data • Outliers • Missing data
![Page 44: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/44.jpg)
Least squares methods
Goal is the find the line that minimizes the residuals.
E =nX
i=1
(yi �mxi � b)2
![Page 45: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/45.jpg)
Least squares methodsE =
nX
i=1
(axi + byi � d)2
Ah = 0 h =
2
4abd
3
5data parameters
Robustness estimator to handle outliers
![Page 46: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/46.jpg)
RANSACRandomized iterative method to fit a parametric model to data based on random sampling of data.
The typical approach 1. Sample the minimum number of points required to fit the model. 2. Fit a model to the sample. 3. Compute inliers. 4. Refine model based on all points.
![Page 47: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/47.jpg)
RANSAC: fitting a circle
1. Select 3 random points. Find center and radius of the circumscribed circle to the triangle formed by the 3 points.
2. Compute distances of all the points to the center of the circle, and calculate inliers and outliers
3. Repeat at most N times
![Page 48: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/48.jpg)
RANSACRandomized iterative method to fit a parametric model to data based on random sampling of data. Pros:• General method suited for a wide range of
model fitting problems • Easy to implement Cons:• Only handles a moderate percentage of
outliers • lots of parameters to tune
![Page 49: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/49.jpg)
Hough TransformA voting scheme to fit a parametric model to data by selecting the model with the most votes.
![Page 50: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/50.jpg)
Hough TransformA voting scheme to fit a parametric model to data by selecting the model with the most votes.
![Page 51: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/51.jpg)
Hough Transform to fit a Circle with known radius R
Each point in the geometric space (left) generates a circle in parametric space (right). The circles in parametric space intersect at the (a,b) that is the center in the geometric space.
y
x a
b
x = a+R cos(✓)
y = b+R sin(✓)
![Page 52: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/52.jpg)
Hough Transform to fit a Multiple Circles with known R
Each point in the geometric space (left) generates a circle in parametric space (right). The circles in parametric space intersect at the (a,b) that is the center in the geometric space.
![Page 53: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/53.jpg)
Hough transformPros:• All points are processed independently, so can handle
occlusions • Some robustness to noise: noisy points unlikely to
contribute consistently to any single bin. • Can detect multipole instances of a model in a single
pass
Cons:• Complexity of search time increases exponentially with
the number of model parameters • Spurious peaks due to uniform noise • Quantization: hard to pick a good grid size.
![Page 54: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/54.jpg)
Detectors and DescriptorsProperties of detectors • Edge detectors (Canny edge detector) • Harris Corner detector • Blob detectors • Difference of Gaussians (DoG)
Properties of descriptors • SIFT • HOG
![Page 55: CS231A Midterm Review - Stanford University...2016/05/06 · CS231A Midterm Review Friday 5/6/2016 Outline • General Logistics • Camera Models • Non-perspective cameras •](https://reader036.vdocuments.site/reader036/viewer/2022071503/6123711236dc7c69ca3e6443/html5/thumbnails/55.jpg)
Good Luck!