Digital Image Processing C++

The dataset is available here.

• Histogram Equalization
• Losspass/Highpass Filter: Mean and Median
• Edge detection: Sobel, Prewitt and Robert
• User-Defined Thresholding

Enviroment

macOS 10.13.6
g++ 4.2.1

Library - only for I/O

opencv 3.4.2

Histogram Equalization

Histogram Equalizaion is a useful image processing method for contrast adjustment.

Command for Compiler

g++ $(pkg-config --cflags --libs opencv) histogram.cpp -o histogram

Command for Execution

./histogram input_img output_img

• For example

./histogram test.tif output.tif

Taking a image, Chapter3, Fig0320(4)(bottom_left).tif, as example.

Losspass/Highpass Filter (3x3)

Spatial Filtering is a useful way for reducing noises of image.

Command for Compiler

g++ $(pkg-config --cflags --libs opencv) filter.cpp -o filter

Command for Execution

./filter input_img

• For example

./filter test.tif

Taking a image, Chapter3, Fig0335(a)(ckt_board_saltpep_prob_pt05).tif, as example.
Note that the size of mean_filter and median_filter is 3x3 for demonstration.

Edge Detection

Here, we mplement three edge detection method – Sobel, Prewitt and Robert.

Command for Compiler

g++ $(pkg-config --cflags --libs opencv) edge_detection.cpp -o edge_detection

Command for Execution

./edge_detection input_img mode

• For example

./edge_detection median_test.tif sobel

Taking a median result above as example.

User-Defined Thresholding

Thresholding is the simplest method of image segmentation.

Command for Compiler

g++ $(pkg-config --cflags --libs opencv) threshold.cpp -o threshold

Command for Execution

./threshold input_file threshold_value

• For example

./threshold test.tif 55

Note that the proper value should be 0≤threshold_value≤255

Taking a image, Chapter3,Fig0316(3)(third_from_top).tif, as example.