You are on page 1of 2
Arab Academy for Science & Technology & Maritime Transport College of Computing & Information Technology Me 7 ad Lecturer: Dr, Aliaa Yousif Course : Pattern Recognition Code : CCS5I6 Marks : 40 Points Date 2 Spring2010 Time: Two hours Final Exam Question I(1Spoints): 1. The two images shown below are quite different, but their histograms are identical. Both images have size 80 x 80, with black (0) and white (1) pixels. a a) Suppose that both images are blurred with a 33 smoothing mask. Would the resultant histograms still be the same? b) Draw the two histograms (before and after smoothing) and explain your answer. Note: the black lines are used to signify the boundaries of the two images but not part of them. . How the "salt & pepper" noise look like? Describe one method that can be used to remove it ? . If you had an image of a photographic negative, suggest a method of producing a ‘positive! of this image . In a given application, an averaging mask is applied to input images to reduce noise and then a Laplacian mask is applied to enhance small details, Would mathematics predict that the result should be the same if the order of the operations were reversed? Perform 3x3 median filtering on the image shown below. Draw a picture of the result. What happens at corners? Letter 'T’ with impulse n ooovo0c00000000 Perform 3 x 3 averaging on the above image of the letter 'T' with impulse noise. Compare the result with that obtained by median filtering. Which would be best if you wanted to recognize the character? MPce\t Question 20 Spoints): 1- From Human Neurons to Artificial Neurons Explain the following \.. Differences in Human Neurons and Artificial Neurons Structure b. Is the following statement true or false? “A multi-layer feed forward network with linear activation functions is more powerful than a single-layer feed forward network with linear activation functions.” Why? c. Discus, why might one want to add noise to the training data when training a neural network? Should the noise be added to the inputs, the target outputs, or both? d. Give brief definitions of supervised learning and unsupervised learning. Describe how each Ky, leaming scheme would be used to train a neural network to perform a categorization task. £ » What is meant by the statement, neural networks are trained, not programmed and explain one major advantage, and one major disadvantage of this fact. Use a simple example to describe pattern recognition processes. 2-a, K-Mean algorithm is a useful clustering algorithm, Draw the flow chart of this algorithm. b, Sample data comes from two classes are given as: Object ‘Atribute 1(X)_| Atribute 2) Object A [1 1 Object B 1 2 [object C13 4 Objet D | 4 5 Suppose that these 4 objects are training data points and each object have 2 attributes. Each attribute = represents coordinate of the object. Determine which objects belong to class 1 and which objects belong to the other class. Question 310 p 1. Covariance is one of the most fundamental numerical measures of the relationship between two variables. What do you conclude from the fowling relationships Pte Relenhip age aoe 2. Principal components analysis (PCA) is a technique that can be used to simplify a dataset, State briefly the steps of this technique and show how it can be used in pattem recognition problems. 3. Ifyou had an image of a face . in color, ic. three monochrome images A; fg; fo for red green and blue. You find that the each tones correspond to’ Fr= 100410; fg= 7046; fb=50#10 ‘Sketch an algorithm that will convert the each color to pure bright white (255,255,255), while leaving other tones untouched. MPCsI2

You might also like