Name: ___________ EE435: Biometric Signal Processing Exam 2 Due Friday, 4/4/14 (Beginning of Class) This is an individual exam. Do not communicate with anyone else: if you have questions, ask me. If you get stuck on a problem, you can ask me for help. In that case, I may determine a point cost for my help, based on the severity of the problem you are having. I will let you know how much the help would cost, then you will have the option to accept the help for a deduction in points, or you can go back and try again on your own. You’ll have this week’s lab period to work on it, but it won’t be due until Friday at the beginning of class. Use any of your notes/graded work/handouts from this semester, the Introduction to Biometrics textbook, the Digital Image Processing using MATLAB textbook (on the biometrics lab bookshelf), or online MATLAB help as needed. When you turn in your exam, provide the following: Hardcopy of: -Your answers to ALL questions (write on these sheets whenever possible). Show your work! -Code for ALL parts of problem 3, including the m-file called prob3b.m -Hand these to me on Friday, 4/4 at the beginning of class. Softcopy of (on the Google Drive in a folder you share with me called “Exam02”): -The image named exam02.png -The commented m-file called prob3b.m for problem 3.b. -Place these on the Google drive by Friday, 4/4 at the beginning of class. Problem Points Possible 1 20 2 30 3 50 Total: 100 Points 1. Suppose there is a new Trusted Traveler program in service at major airports in the European Union. For an initial fee of €50, plus an annual fee of €99, a person can bypass standard airport security and use the speedy check-in lane used by the flight crews. The initial fee covers enrollment and a criminal background check. Once enrolled, and assuming their annual payment has been received, a trusted traveler simply walks up to an iris scanner, presents their iris and when recognized, may make their way to their plane. If for some reason they cannot be recognized, they go to a friendly customer service center to figure out why not. Suppose that since its creation in March 2011, the following data was collected: -38,154 persons paid their initial fees, and successfully enrolled in the system. -Among Trusted Travelers, there were 1,355,221 attempts to use their status to bypass normal security. -10,640 persons who were not Trusted Travelers attempted to bypass normal security (because they either didn’t understand that you had to pay and be enrolled, or were news reporters testing security, or were bad guys) -There were 8 patrons who attempted to but was unable to participate because their irises were not able to be captured (i.e, they were blind or had other iris diseases or conditions). Also, two of the Three Stooges attempted to enroll but were unable because someone had poked one of their eyes out. Of these, 1,352,914 Trusted Travelers bypassed normal security w/o problem. The remainder of the Trusted Travelers could not enter, and had to go to customer service to fix it. In addition, 1 person managed to get through the system that was not a Trusted Traveler. Using this information, answer the following: a. (4 pts) Is using iris in this way identification or verification? Why? b. (4 pt) How many FTEs occurred? What is the FTE rate (in %)? c. (4 pts) What is the FAR (in %)? d. (4 pts) What is the FRR (in %)? e. (4 pts) Based on your answers to parts (c) and (d), would you say that The Trusted Traveler system was geared more towards security or convenience? Why? 2. Your senior design project is based on gait recognition, where a gait system will be tied to access to the biometrics lab. You are building a system that is supposed to tell which of three midshipmen is entering the lab based on their gait as they walk up to the door. You’ve studied three classmates, labeled A, B, and C, their gait characteristics, and how often they go into the lab. Gate is measured only using walking speed (fast, medium or slow) as a feature. You’ve taken some measurements of mids A, B and C over the course of the semester and determined some a priori knowledge about this situation. When reviewing the data you’ve collected about these three mids and their lab usage, of all their lab entries, A enters the lab 15% of the time, B enters 25% of the time, and C enters 60% of the time. You’ve also taken some measurements of their gait characteristics and have come up with the following information: -Midshipman A walks fast 80% of the time and medium 20% of the time. -Midshipman B walks fast 20% of the time, medium 50% of the time, and slow 30% of the time. -Midshipman C walks fast 10% of the time, medium 20% of the time and slow 70% of the time. a. (12 pts) Fill in the probability tree below for this pattern recognition problem. Include probabilitistic expressions such as p(fast|C) as well as numbers, so I can follow your tree’s construction. If needed, use another sheet of paper for calculations. b. (3 pts) What is the probability that the person writing on the pad is walking fast? Medium? Slow? c. (3 pts) What is the probability that the midshipman entering the lab is mid B? Why? d. (6 pts) Your system decides which midshipman it is just by how fast they walk up to the lab door. State your overall system decision rule by filling in the identity of the midshipman below (show work): If the mid is walking fast, it must be _____________________. If the mid is walking medium, it must be _____________________. If the mid is walking slowly, it must be ____________________. e. (3 pts) Describe how well this system works if there were 4 mids that were being recognized, vice 3 mids, but the same gait speeds. f. (3 pts) When do you expect to have the smallest error percentage; if the person walks fast, medium or slow? Why? 3. Download the image 0024_L_0015.bmp from the course website. This is a high-resolution nearinfrared image of an iris from the University of Bath iris image database. Use MATLAB to find your answers. a. (20 pts total) Process this image using the following steps, in order, and answer the questions. Take this original image, invert it, square the values, and threshold it with a threshold value of “mean + 2 times its standard deviation”. Do not perform any further morphology. When you do this: (1) How many objects are there in the resulting binary image? _____________ (2) What object number corresponds to the object with the largest area? _____________ (3) What is the area of the largest object (from the previous question)? _____________ (4) What is the perimeter of the largest object? _____________ Note: unnecessary looping will be penalized! You do NOT need to loop to calculate areas or perimeters. b. (20 pts total) Take the original image, and create MATLAB code that will automatically find the approximate area of the glare spot, using morphology and/or other image processing techniques you’ve learned. Note: this is a script file, NOT a function. You could possibly make necessary changes to the procedure in part (a) that results in a binary image with only one object (i.e., the glare spot), then use that one object to automatically find the area of the glare spot. Put your code in a single m-file called prob3b.m that you will provide to me in softcopy, and I will run it myself. If you wrote any other functions besides built-in MATLAB functions that prob3b.m uses, be sure to include them too (for this problem, you don’t need to create any special functions, but you can if you wish). Be sure to put COMMENTS in the code so I can follow your thought process. The final line in this prob3b.m m-file should be the following, so I will know the value of area that you compute: disp(sprint(‘The area of the glare spot is %d pixels\n’, x)) where x is the area your program computes. Note: unnecessary looping will be penalized! You do NOT need to loop to calculate area. In the space below, briefly describe your how you went about doing this (be more general than the comments in your code). c. (10 pts total) Take your process from part (b) and use it to create a color iris image where the glare spot is completely filled in with bright red (the rest of the image should be shades of gray). Using imwrite, write the color image into a file called “exam02.png”, and place it in your Exam02 folder on the Google Drive. In the space below, briefly describe how you went about doing this.