Suppose we have a video of two white points moving toward each other at a uniform rate against a black background. We encode it via MPEG. In one I frame the two points are 100 pixels apart; in the next I frame they have merged. The final point of merger happens to lie at the center of a 16 x 16 macroblock. (a) Describe how you might optimally encode the Y component of the intervening B (or P) frames. (b) Now suppose the points are in color and that the color changes slowly as the points move. Describe what the encoding of the U and V values might look like.

Suppose we have a video of two white points moving toward each other at a uniform rate against a black background. We encode it via MPEG. In one I frame the two points are 100 pixels apart; in the next I frame they have merged. The final point of merger happens to lie at the center of a 16 x 16 macroblock. (a) Describe how you might optimally encode the Y component of the intervening B (or P) frames. (b) Now suppose the points are in color and that the color changes slowly as the points move. Describe what the encoding of the U and V values might look like.

Database System Concepts

7th Edition

ISBN:9780078022159

Author:Abraham Silberschatz Professor, Henry F. Korth, S. Sudarshan

Publisher:Abraham Silberschatz Professor, Henry F. Korth, S. Sudarshan

Chapter1: Introduction

Section: Chapter Questions

Problem 1PE

See similar textbooks

Similar questions

Consider a color image with 6x8 pixels. R, G, and B components are stored with 8 bits. There are 2^n distinct colors in the image and each pixel is stored with n bits in index color format. What is the MAXIMUM value of n that makes index color format more advantageous than RGB color format?
We are building a CNN network for image classification. The image dmension is 64*64*3. Here is the network. CONV1: 64 filters of size 3*3*x1, stride 1, padding p1, (output image dimension is same as input); CONV2: 32 filters of size 5*5*x2, stride 1, padding p2, (output image dimension is same as input); POOL1: filter size 2*2, stride 2, padding 0 (output dimension is D) a) Find, D, x1, x2, p1, p2. (10) b) How many parameters this model needs to learn for this problem.
Make a coding using OpenCV to process the image convolution that follows question no. 4. a)-e). Use chess board 8x8. Use operator Kernel with size 3*3. a) original image: Chess Board 8x b) Image captures for those pictures a) original image Who are they? d) Negative b) Mean e) Low pass filter, threshold 20 c) Mid-Point f) Weighted average
Construct 5 × 5 noisy image matrix from your Student ID as discussed below?Example: Assume your Student ID is 437818854, and then use the first five numbers as rows and thelast five numbers as columns. Use the following formula to fill each pixel value g(x,y) = Row × Col + 12
Integral image is obtained by summing all the pixels before each pixel (Naively you can think of this as similar to the cumulative distribution function where a particular value is obtained by summing all the values before). Let's take an example to understand this. Suppose we have a 5x5 binary image as shown below. The integral image is shown on the right. 2, 1, 2, 2], 1, 1, 3, 3], 5, 5], 8], [1, 2, 2, 5, [2, 4, 4, 8, [ 3, 5, 5, 9, 10]]. Integral image All the pixels in the integral image are obtained by summing all the previous pixels. Previous here means all the pixels above and to the left of that pixel (inclusive of that pixel). For instance, the 3 (blue circle) is obtained by adding that pixel with the above and left pixels in the input image i.e. 1+0+0+1+0+0+0+1 = 3. [[1, 0, 0, 1, 0], [0, 0, 0, 1, 0], [0, 1, 0, 1, 0], [1, 1, 0, 1, 0], [1, 0, 0, 0, 1]] Original image [[1, 1, 1, [1, Write a function integral_image () that takes as parameter an image in the form of a nested list A…
Take a 256 × 256 grayscaled image. Hide a (text) message of length 15 to 25 in the image such that at the pixel ([p/8], [q/8]). the first bit of the hidden message is positioned P = 839 9 = 283
Consider an image strip of size 50 X 100. It consists of five vertical stripes. The grey levels of the stripes from left to right are 128,64,32,16 and 8 pixels. The corresponding widths of the stripes are 35, 30, 20, 10 and 5 pixels respectively. Apply run-length encoding on the image strip and compute the efficiency of the run-length coding.
Problem 1. Download the file Square.tif and answer the following questions: i. Smooth the image with a 3x3 spatial filter. ii. Smooth the image with an implementation of Gaussian filter in frequency domain. Change the bandwidth of the filter and show how these changes affect the result. iii. Visualize that filtering in frequency domain can cause wraparound error in spatial domain and explain how can this interference be avoided. Apply your idea as a preprocessing step in part (ii) and display your results. iv. Sum up your conclusions in part (ii) and (iii), and List the Basic Steps required in DFT filtering. Implement all these steps in an M-file, named YOUfft2 (replace YOU with your name). This file must get an arbitrary image as input, smooth the image with a Gaussian filter, and give the filtered image in output. The process of filtering should be done in frequency domain. Problem 2. Download the file CA03.jpg. i.Add Gaussian noise N( z =0, o =10/256) to the image and then apply…
6. Why aren't B-frames used as reference frames for motion compensation? Suppose there is a mode where any frame type can be specified as a reference frame. Discuss the tradeoffs of using reference B-frames instead of P-frames in a video sequence (i.e., eliminating P-frames completely).
6. Suppose we have an image with 40x42x3 dimensions, and we apply a convolution operation with six 5x5 filters, what are the dimensions of the output map?
1. MPEG output consists of three kinds of frames – I, P, and B frames. For scenes in which the camera and background are stationary and one or two actors are moving around slowly, nearly all the pixels will be identical from frame to frame. Here, just subtracting each frame from the previous one and running JPEG on the difference would do fine. However, for scenes where the camera is panning or zooming, this technique fails badly. Let us assume that the background of scenes is stationary, and the scenes are sent on UDP. Due to an unreliable communication, UDP may loss some of the frames. Which frame (I, P or B) is significant? Thank you so much in advance.
When embedding a 10-bit message in an image using LSB replacement, write down the probability that the embedding will introduce k ∈ {0,1,...,10} embedding changes.