Object Removal in Multi-View Photos Aaron McClennon-Sowchuk, Michail Greshischev.

Slides:



Advertisements
Similar presentations
Distinctive Image Features from Scale-Invariant Keypoints
Advertisements

Distinctive Image Features from Scale-Invariant Keypoints David Lowe.
Object Recognition from Local Scale-Invariant Features David G. Lowe Presented by Ashley L. Kapron.
The fundamental matrix F
Lecture 11: Two-view geometry
CSE473/573 – Stereo and Multiple View Geometry
QR Code Recognition Based On Image Processing
MASKS © 2004 Invitation to 3D vision Lecture 7 Step-by-Step Model Buidling.
Object Recognition using Invariant Local Features Applications l Mobile robots, driver assistance l Cell phone location or object recognition l Panoramas,
Distinctive Image Features from Scale- Invariant Keypoints Mohammad-Amin Ahantab Technische Universität München, Germany.
Image alignment Image from
Instructor: Mircea Nicolescu Lecture 13 CS 485 / 685 Computer Vision.
IBBT – Ugent – Telin – IPI Dimitri Van Cauwelaert A study of the 2D - SIFT algorithm Dimitri Van Cauwelaert.
Object Recognition with Invariant Features n Definition: Identify objects or scenes and determine their pose and model parameters n Applications l Industrial.
Robust and large-scale alignment Image from
Object Recognition with Invariant Features n Definition: Identify objects or scenes and determine their pose and model parameters n Applications l Industrial.
Distinctive Image Feature from Scale-Invariant KeyPoints
Distinctive image features from scale-invariant keypoints. David G. Lowe, Int. Journal of Computer Vision, 60, 2 (2004), pp Presented by: Shalomi.
Scale Invariant Feature Transform (SIFT)
1 Invariant Local Feature for Object Recognition Presented by Wyman 2/05/2006.
CS4670: Computer Vision Kavita Bala Lecture 8: Scale invariance.
Sebastian Thrun CS223B Computer Vision, Winter Stanford CS223B Computer Vision, Winter 2005 Lecture 3 Advanced Features Sebastian Thrun, Stanford.
Scale-Invariant Feature Transform (SIFT) Jinxiang Chai.
Automatic Camera Calibration
CSE 185 Introduction to Computer Vision
Distinctive Image Features from Scale-Invariant Keypoints By David G. Lowe, University of British Columbia Presented by: Tim Havinga, Joël van Neerbos.
Computer vision.
Final Exam Review CS485/685 Computer Vision Prof. Bebis.
By Yevgeny Yusepovsky & Diana Tsamalashvili the supervisor: Arie Nakhmani 08/07/2010 1Control and Robotics Labaratory.
Internet-scale Imagery for Graphics and Vision James Hays cs195g Computational Photography Brown University, Spring 2010.
1 Interest Operators Harris Corner Detector: the first and most basic interest operator Kadir Entropy Detector and its use in object recognition SIFT interest.
The Brightness Constraint
Overview Harris interest points Comparing interest points (SSD, ZNCC, SIFT) Scale & affine invariant interest points Evaluation and comparison of different.
CS654: Digital Image Analysis Lecture 8: Stereo Imaging.
1 Formation et Analyse d’Images Session 7 Daniela Hall 25 November 2004.
CSCE 643 Computer Vision: Extractions of Image Features Jinxiang Chai.
Computer Vision : CISC 4/689 Going Back a little Cameras.ppt.
Wenqi Zhu 3D Reconstruction From Multiple Views Based on Scale-Invariant Feature Transform.
Lecture 7: Features Part 2 CS4670/5670: Computer Vision Noah Snavely.
Distinctive Image Features from Scale-Invariant Keypoints Ronnie Bajwa Sameer Pawar * * Adapted from slides found online by Michael Kowalski, Lehigh University.
Distinctive Image Features from Scale-Invariant Keypoints David Lowe Presented by Tony X. Han March 11, 2008.
Jack Pinches INFO410 & INFO350 S INFORMATION SCIENCE Computer Vision I.
Feature Matching. Feature Space Outlier Rejection.
CSE 185 Introduction to Computer Vision Feature Matching.
A Tutorial on using SIFT Presented by Jimmy Huff (Slightly modified by Josiah Yoder for Winter )
October 1, 2013Computer Vision Lecture 9: From Edges to Contours 1 Canny Edge Detector However, usually there will still be noise in the array E[i, j],
Here today. Gone Tomorrow Aaron McClennon-Sowchuk, Michail Greshischev.
776 Computer Vision Jan-Michael Frahm Spring 2012.
Instructor: Mircea Nicolescu Lecture 5 CS 485 / 685 Computer Vision.
Linearizing (assuming small (u,v)): Brightness Constancy Equation: The Brightness Constraint Where:),(),(yxJyxII t  Each pixel provides 1 equation in.
Recognizing specific objects Matching with SIFT Original suggestion Lowe, 1999,2004.
CSCI 631 – Foundations of Computer Vision March 15, 2016 Ashwini Imran Image Stitching.
Blob detection.
SIFT.
776 Computer Vision Jan-Michael Frahm Spring 2012.
SIFT Scale-Invariant Feature Transform David Lowe
Interest Points EE/CSE 576 Linda Shapiro.
Lecture 07 13/12/2011 Shai Avidan הבהרה: החומר המחייב הוא החומר הנלמד בכיתה ולא זה המופיע / לא מופיע במצגת.
Distinctive Image Features from Scale-Invariant Keypoints
Project 1: hybrid images
CS4670 / 5670: Computer Vision Kavita Bala Lec 27: Stereo.
Feature description and matching
CSE 455 – Guest Lectures 3 lectures Contact Interest points 1
From a presentation by Jimmy Huff Modified by Josiah Yoder
The SIFT (Scale Invariant Feature Transform) Detector and Descriptor
SIFT keypoint detection
SIFT.
Feature descriptors and matching
Presented by Xu Miao April 20, 2005
Presentation transcript:

Object Removal in Multi-View Photos Aaron McClennon-Sowchuk, Michail Greshischev

Objectives  Remove an object from a set of images by using information (pixels) from other images in the set.  The images must be of the same scene but can vary in time of taken and/or perspective of scene.  The allowed variance in time means objects may change location from one image to the next. Applications: stock photography, video surveillance, etc.

Steps 1. Read Images 2. Project images in same perspective 3. Align the images 4. Identify differences 5. Infill objects

Reading Images  How are images represented? –Matrices (M x N x P) –M is the width of the image –N is the height of the image –P is 1 or 3 depend on quality of image 1: binary (strictly black/white) or gray-scale images 3: coloured images (3 components of colour: R,G,B)  What tools are capable of processing images? –Many to choose from but MatLab is ideal for matrices. –Hence the name Mat(rix) Lab(oratory)

Object Removal in Multi-View Photos Image Rectification

Figure 1: Example rectification of source images (1) to common image plane (2). 1  Transformation process used to project two-or-more images onto a common image plane.  Corrects image distortion by transforming the image into a standard coordinate system. 1

Image Rectification To perform a transform...  Cameras are calibrated and provide internal parameters resulting in an essential matrix representing relationship between the cameras. –We don’t have access to camera’s internal parameters. –What if single camera was used?  The more general case (without camera calibration) is represented by the fundamental matrix. 2

Fundamental Matrix  Algebraic representation of epipolar geometry.  3×3 matrix which relates corresponding points in stereo images.  7 degrees of freedom, therefore at least 7 correspondences are required to compute the fundamental matrix. 3

Corresponding Points  Figure out which parts of an image correspond to which parts of another image. –But what is a ‘part’ of an image?  ‘part’ of an image is a Spatial Feature.  Spatial Feature Detection is the process of identifying spatial features in images.

Spatial Feature Detection - Edges  Canny, Prewitt, Sobel, Difference of Gaussians... Figure 2: Example application of Canny Edge Detection 4

Spatial Feature Detection - Corners  Harris, FAST, SUSAN Figure 2: Example application of Harris Corner Detection 5

Feature Description  Simply identifying a feature point is not in itself useful. – consider how one would attempt to match detected feature points between multiple images.  Scale-invariant feature transform (SIFT) offers robust feature description. 6 –Invariant to scale –Invariant to orientation –partially invariant to illumination changes

SIFT  Uses Difference of Gaussians along with multiple smoothing and resampling filters to detect key points (Feature Points with descriptor data)  Key point specifies 2D location, scale, and orientation.

SIFT Figure 3: Sample image for SIFT application. 7

SIFT – Feature Points Figure 4: Detected feature points via SIFT. 7

SIFT – Key Point Figure 5: A SIFT key point in detail. 7

SIFT - Matching  Matches key points by identifying nearest neighbour with the minimum Euclidean distance.  Ensures robustness via...  Cluster identification by Hough transform voting.  Model verification by linear least squares.

SIFT - Matching Figure 5: Example of matched SIFT key points. Note its tolerance to image scale and rotation.

SIFT – Suitable for Multi-View?  SIFT fails to accurately match key points between images which vary significantly in perspective. Figure 7 & 8: Comparison of SIFT accuracy with varying perspective angles. Left image is 45 degrees with 152 matches. Right image is 75 degrees with 11 matches. 8

SIFT – Suitable for Multi-View?  SIFT fails to accurately match key points between images which undergo non-scalable affine transformation or projection. Figure 9: SIFT fails to identify any key point matches between rotated images on a cylinder. 8

ASIFT  Affine-SIFT (ASIFT) is a new framework for fully affine invariant image comparison.  Uses existing SIFT key point descriptors, but matching algorithm has improved.

ASIFT – Improvements over SIFT  Simulated images are compared by a rotation, translation and zoom-invariant algorithm. –(SIFT normalizes translation and rotation and simulates zoom.)

ASIFT – Improvements over SIFT Figure 10: ASIFT (left) identifies 165 matches compared to SIFT’s (right) 11 matches on surface rotated 75 degrees. 8

ASIFT – Improvements over SIFT Figure 10: ASIFT identifies 381 matches between rotated surfaces. 8

Image Rectification  Quick Review... 1.Given multiple images of the same scene from different perspectives... 2.We have identified & matched feature points using ASIFT.  We now have sufficient matching points to calculate the fundamental matrix.

Calculating Fundamental Matrix  Random Sample Consensus (RANSAC) is used to eliminate outliers from matched points. 1.Select 7 points at random. 2.Use them to compute a Fundamental Matrix between the image pair. 3.Project every point in the dataset onto the conjugate image pair using the Fundamental Matrix. 4.If at least 7 points were projected closer to their actual locations than their allowable errors, stop. 5.Use those 7 points to calculate final Fundamental Matrix.

Example Image Rectification  Input Images

Example Image Rectification  ASIFT Matches

Example Image Rectification  RANSAC selection

Example Image Rectification  Resulting Rectification

Identifying Image Differences  Possible Methods: 1. Direct subtraction 2. Structural Similarity Index (SSIM) 3. Complex Waveform SSIM

Identifying Image Differences 1. Direct subtraction –Too good to be true!(way too much noise)

Identifying differences 2. Structural Similarity Index (SSIM) –Number 0-1 indicating how “similar” two pixels are. –1 indicates perfect match, 0 indicates no similarities at all –Number calculated based on: – Luminance, function of the mean intensity for gray-scale image – Contrast, function of std.dev of intensity for gray-scale image

Object Removal in Multi-View Photos Complex Waveform SSIM

 SSIM vs Complex Waveform SSIM(CWSSIM) SSIMCWSSIM Sensitive to pixel shifting (Spatial Shifts) Tolerates small amounts of pixel shifting (Spatial Shifts) Equally weight given to low resolution and high resolution differences. Bands are scalable. Reports incorrect magnitude of error in blurred images. Correctly identifies level of error in blur.

CWSSIM - Implementation  Steerable Pyramid constructed for each image –(Steerable Pyramid is a linear multi-scale, multi- orientation image decomposition)  SSIM value calculated for each band, from high to low frequency.  SSIM values for each band are scaled and summed.

CWSSIM - Implementation  Analyzing bands with multi-scale, multi-orientation image decomposition instead of direct pixel comparison provides tolerance for Spatial Shifts.  By reducing the contribution SSIM indexes belonging to high frequency bands we can reduce noise. –…but we lose recognition of changes in that frequency.

CWSSIM - Example  Input Images

CWSSIM - Example  Application of CWSSIM with equal frequency weights.

CWSSIM - Example  Input Images

CWSSIM - Example  Application of CWSSIM with decreased low frequency weight.

Identifying differences  Once again, way too much noise.  SSIM map: 0  black pixel1  white pixel

–Concerns: –Identify regions to copy Calculate a bounding box (smallest area surrounding entire blob) –How to distinguish noise from actual objects? Area - those blobs with area below threshold are ignored location - those blobs along an edge of image are ignored. –Copying method Direct – images from same perspectives Manipulated pixels – images from different perspectives. Infilling the objects

 Original bounding box results: Matlab returns Left position Top position Width and Height of each box

Infilling the objects  Result with small blobs and blobs along edges ignored:  Left: 119  Top: 52  Width: 122  Height: 264

Infilling the objects  Once regions identified, how can pixels be copied? –Same perspective – direct copy is possible.

Infilling the objects  Result of direct copying

Infilling the objects  Different perspectives –Goal: remove black trophy from left image

Infilling the objects  Direct copying produces horrendous results! Rectified image Result

Work to come...  Copying techniques –Need better method for infilling objects between images in different perspectives. Perhaps use same alignment matrix.  Anti-Aliasing –Method to smooth the edges around pixels copied from one image to another – example looks alright but could improve other test cases  User friendly interface –Current state: a dozen different MatLab scripts. –In the perfect world, we’d have a nice interface to let user load images and clearly displa

Thank You! Questions?

References 1. Oram, Daniel (2001). "Rectification for Any Epipolar Geometry“ 2. Fusiello, Andrea ( ). "Epipolar Rectification" Richard Hartley and Andrew Zisserman (2004). “Multiple View Geometry in Computer Vision Second Edition” 4. Ma,Yi. (1996) Basic Image Processing Demos (for EECS20) 5. Mark Nixon & Alberto Aguado (2002), Feature Extraction & Image Processing, Newnes 6. Lowe, D. G., “Distinctive Image Features from Scale-Invariant Keypoints”, International Journal of Computer Vision, 60, 2, pp , Andrea Vedaldi and Brian Fulkerson (2005), “VL_SIFT” 8. Jean-Michel Morel and Guoshen Yu (2010), “SIFT and ASIFT”, ASIFT: A New Framework for Fully Affine Invariant Image Comparison

References  Z. Wang and A. C. Bovik, “Image quality assessment: from error visibility to structural similarity,” IEEE Trans. Image Processing, vol. 13, pp. 600 – 612, Apr html html