Assignment 1: Algorithms and Systems

 Algorithms and Systems

Assignment One (20%)

Objectives
x Getting familiar with OpenCV 3.2.0
x Reading, processing and displaying images using OpenCV 3.2.0
x Understanding the principles of single sensor digital cameras
Task One (15%)
Images captured by semiconductor sensors have to be processed before they are passed to later
stages in computer vision systems. In this assignment, you are required to develop a C/C++
program that implements the color image processing chain converting raw image data
captured by a CMOS image sensor into true color RGB images. The chain consists of three
components: color interpolation, color correction and gamma correction as shown below.
You are provided with two bmp files test1.bmp and test2.bmp, which contain the raw
images (Bayer Pattern Color Filter Array data) directly captured by a CMOS image sensor. The

images have resolution 640x480 pixels. The Bayer pattern used in the CMOS sensor is shown
below:
G R G R …
B G B G …
G R G R
B G B G
…
Color Interpolation - You are required to implement the bilinear color interpolation algorithm to
produce an RGB image with the same width and height as the input raw image. Therefore, you need to
take into account boundary conditions, as pixels around the image boundary may not have neighboring
pixels on one or two sides. Try to find the most efficient and simple solution to interpolate boundary
pixels.
Color Correction - Color Correction is a matrix operation. You should use the following matrix that
was optimized for this CMOS image sensor:
»
»
»
¼
º
«
«
«
¬
ª
»
»
»
¼
º
«
«
«
¬
ª
 
 
 

»
»
»
¼
º
«
«
«
¬
ª
B
G
R
B
G
R
0.18 0.44 1.62
0.24 1.29 0.05
1.18 0.05 0.13
'
'
'
Matrix multiplication can result in output values <0 or 255. Thus, you need to check the values and
clip them to 0 or 255 if needed.
Gamma Correction - Gamma correction should be implemented using a look-up table. The table
entries must be calculated using J = 0.5 and implement the transfer function to achieve Gamma
correction. You need to find the expression how to calculate and fill the table.
The program should display the raw image (data), color interpolated image, color corrected image
and the gamma corrected image in a single window as illustrated below.
Raw Data Color Interpolated
Image
Color Corrected
Image
Gamma Corrected
Image

Requirements on coding
1. The program should be named as “colorChain” and shall take a BMP image file as the input
image, e.g. colorChain bmpfile
2. No other third-party libraries should be used in the program except OpenCV 3.2.0. The code
has to be in C/C++.
3. The code should be modularized with detail comments AND all source code should be
placed in a single file “colorChain.cpp” or “colorChain.c”.
Marking Scheme
1. Zero marks may be graded if your code cannot be compiled.
2. Program structure, comments and usability (3%)
3. Display of the raw image (3%)
4. Generation and display of the color interpolated image (3%)
5. Generation and display of the color corrected image (3%)
6. Generation and display of the gamma corrected image (3%)
Task Two (5%)
You also need to produce a report of no more than one page (report.txt) which covers the
following
1. A brief description of your implementation of the color processing chain on how to deal
with boundary pixels, your observations and discussions on the quality of the generated
images at each stage.
2. Comparison of the color-corrected and gamma-corrected images generated by running your
program on test3.bmp, discussion on the differences from test3.jpg and possible
causes of the differences. Notice that the Bayer raw data file, test3.bmp, is simulated
from the image test3.jpg by sampling the RGB data with the same CFA Bayer pattern
as shown above.
3. Comments and discussion on the spectral characteristics of the lighting sources under which
images “light01.jpg” and “light02.jpg” were taken. The images were taken by the
same camera with the exactly same settings. You may use sunlight as a reference for
comparison.
Submission
Zip the SOURCE file and report.txt to your_login_name.zip. The zip file has to be submitted
in Moodle
IMPORTANT:
a) DO NOT include and submit any object files and images in the zip file. Your submission may
not be accepted if you do so.