See Through Fog Imaging Project: P06441

Slides:



Advertisements
Similar presentations
Warm up 1. Solve 2. Solve 3. Decompose to partial fractions -1/2, 1
Advertisements

Grey Level Enhancement Contrast stretching Linear mapping Non-linear mapping Efficient implementation of mapping algorithms Design of classes to support.
1 MA 1128: Lecture 19 – 4/19/11 Quadratic Formula Solving Equations with Graphs.
Agricultural and Biological Statistics
1 Chapter 1: Sampling and Descriptive Statistics.
CONTOUR LINES.
Adaptive Rao-Blackwellized Particle Filter and It’s Evaluation for Tracking in Surveillance Xinyu Xu and Baoxin Li, Senior Member, IEEE.
Objectives Learn how to implement the sequential search algorithm Explore how to sort an array using the selection sort algorithm Learn how to implement.
Probability Distributions Continuous Random Variables.
 2006 Pearson Education, Inc. All rights reserved Searching and Sorting.
RATIOS OF SCALE DRAWINGS. SCALE DRAWINGS SCALE DRAWINGS: A scale drawing is a drawing that represents a real object. The scale of the drawing is the ratio.
Computer Vision - A Modern Approach Set: Segmentation Slides by D.A. Forsyth Segmentation and Grouping Motivation: not information is evidence Obtain a.
Navigating and Browsing 3D Models in 3DLIB Hesham Anan, Kurt Maly, Mohammad Zubair Computer Science Dept. Old Dominion University, Norfolk, VA, (anan,
Abstract Design Considerations and Future Plans In this project we focus on integrating sensors into a small electrical vehicle to enable it to navigate.
FEATURE EXTRACTION FOR JAVA CHARACTER RECOGNITION Rudy Adipranata, Liliana, Meiliana Indrawijaya, Gregorius Satia Budhi Informatics Department, Petra Christian.
Chapter 3: Central Tendency. Central Tendency In general terms, central tendency is a statistical measure that determines a single value that accurately.
Eva Math Team 6 th -7 th Grade Math Pre-Algebra. 1. Relate and apply concepts associated with integers (real number line, additive inverse, absolute value,
COURSE: JUST 3900 INTRODUCTORY STATISTICS FOR CRIMINAL JUSTICE Instructor: Dr. John J. Kerbs, Associate Professor Joint Ph.D. in Social Work and Sociology.
Additional Problems.
Students: Anurag Anjaria, Charles Hansen, Jin Bai, Mai Kanchanabal Professors: Dr. Edward J. Delp, Dr. Yung-Hsiang Lu CAM 2 Continuous Analysis of Many.
Optimising Cuts for HLT George Talbot Supervisor: Stewart Martin-Haugh.
Problem Solving Engineering Technology Mr. Austin.
Lecture 7 - Systems of Equations CVEN 302 June 17, 2002.
Centroids part 2 Getting rid of outliers and sorting.
Comparison of Digital Image Filtering Techniques Kevin Liu Thomas Jefferson High School for Science and Technology.
1 Psych 5500/6500 Measures of Variability Fall, 2008.
Measures of variability: understanding the complexity of natural phenomena.
UNIT 5.  The related activities of sorting, searching and merging are central to many computer applications.  Sorting and merging provide us with a.
Section P.7  An expression is an algebraic statement that does not have an “=“  An equation is two expressions joined by an “=“
ALGORITHMS.
Solving Equations Containing First, we will look at solving these problems algebraically. Here is an example that we will do together using two different.
CfE Advanced Higher Physics
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.
Chapter 6: Descriptive Statistics. Learning Objectives Describe statistical measures used in descriptive statistics Compute measures of central tendency.
Computer Graphics CC416 Lecture 04: Bresenham Line Algorithm & Mid-point circle algorithm Dr. Manal Helal – Fall 2014.
Radical Functions and Equations ( 무리함수 ) Radical sign index The term is called a radical. The symbol is called a radical sign(sometimes just called radical.
IMAGE PROCESSING APPLIED TO TRAFFIC QUEUE DETECTION ALGORITHM.
Exploratory Data Analysis
Image Processing Methods Using MATLAB
19 Searching and Sorting.
Different Types of Data
4.5: Linear Approximations, Differentials and Newton’s Method
Dr. Jim Rowan ITEC 2110 Video Part 2
HW: Worksheet Aim: How do we solve fractional equation?
12.1 Acceleration.
Mathematical Modeling and Variation 1.10
Statistics A statistic: is any number that describes a characteristic of a sample’s scores on a measure. Examples are but not limited to average (arithmetic.
Simulation-Based Approach for Comparing Two Means
Multidisciplinary Engineering Senior Design Project P06441 See Through Fog Imaging Preliminary Design Review 05/19/06 Project Sponsor: Dr. Rao Team Members:
Solving Rational Equations
M3U5D5 Warmup Simplify the expression.
Guess the Age Mat 075 Fall 2016.
Introduction to Summary Statistics
Regression and Residual Plots
Vision processing for robot navigation
Sensors Training.
Dr. Jim Rowan ITEC 2110 Video Part 2
Introduction to Summary Statistics
Solving Equations Containing
Larry Braile, Purdue University
Project P06441: See Through Fog Imaging
Functions and Their Graphs
Sampling Distributions
Chapter 6 Section 5.
Section P7 Equations.
Chapter 4 Motion.
Chapter 1 Motion.
Chapter 3: Central Tendency
Image Enhancement in Spatial Domain: Point Processing
Presentation transcript:

See Through Fog Imaging Project: P06441 Fog is a natural occurring phenomenon which inhibits visibility. Fog has effects on all facets of transportation whether it be flying over or through it in an airplane, or driving through it with a car. It also affects the visibility of images through the use of cameras in aerial photography. By eliminating the fog from these images, a number of exciting applications become possible. Using video cameras in transportation will allow the vehicle operator to have a clear view of the surrounding terrain through the use of video sampling and a LCD screen. Using a similar process in aerial photography will allow video reconnaissance of foggy areas. Two algorithms have been developed to successfully remove fog from images. Each algorithm uses the same basic equation to defog the image: Equation: Algorithm One Algorithm Two Algorithm One uses the above equation to defog the images by making an initial guess as to the values of the two main variables, B and C0, and scales the pixels within the image to be between zero and one. It then runs the algorithm using the initial guesses and searches for the smallest pixel. Once this is found, it then varies B until this pixel value reaches zero. It then locates the largest pixel and varies C0 until the pixel value becomes one. It then checks to see if the difference between the new and old B and C0 values are less then a threshold value, a very small number somewhere around 1E10^-15. If it is, then the B and C0 values have been found and the image can be defogged, if not, then repeats the process using the calculated B and C0 values. Examples can be seen below. The images below are an example of how the defogging algorithm was applied to an image. The first image is a common picture of a baboon and the graph below is its histogram. The middle image is the same image after fog was induced artificially induced upon, using greater “distance” for pixels that are nearer the top of the image. The histogram below this image shows a histogram that is substantially different from the original histogram. The third image is the image after the second defogging algorithm has been applied to it. It is visually identical to the original image. The histogram is also nearly identical to the original histogram, however, its right side has been stretched to the right (toward one) a small, but noticeable, amount (by .1 - .2). Original Image Fog Induced Image Defogged Image Original Histogram Fogged Histogram Defogged Histogram Algorithm Comparison The Algorithms were compared in order to get a better understanding of how they may be used, be it for surveillance or traveling purposes. Algorithm Two defogs faster then Algorithm One in most cases Algorithm Two has a smaller Root Mean Square Error (the resulting image is clearer) Algorithm Two tends to determine a B and C0 which is closer to the actual B and C0 It can be concluded that Algorithm Two is the better of the two algorithms Future Work Design a system to retrieve distance information Devise a method to apply these techniques to color images Implement a video system capable of removing fog in real-time Find a way to apply the algorithm to images where B is not constant Team Members Philip Edwards, Computer Engineering William Parsons, Computer Engineering Acknowledgments Project Sponsor: Dr. Raghuveer Rao Project Mentor: Professor George Slack