Image File Processing

Modern C/C++ BMP image-processing desktop application built with EasyX.

---

Overview

Image File Processing is a desktop image-processing application developed in C/C++ using the EasyX graphics library.

The program reads a 24-bit uncompressed BMP image, displays it through a modern dark graphical interface, and applies multiple image-processing operations such as grayscale conversion, Sobel edge detection, sharpening, salt-and-pepper noise generation, and median filtering.

This project demonstrates practical skills in:

• Binary BMP file processing
• Pixel-level image manipulation
• C/C++ memory management
• Convolution-based image algorithms
• Modular software design
• EasyX graphical interface development
• Debugging of low-level image format issues

---

Key Features

Feature	Description
BMP Loader	Reads and validates 24-bit uncompressed BMP files
Image Preview	Displays the image inside a scaled preview area
Grayscale Conversion	Converts BGR image data into grayscale intensity values
Sobel Edge Detection	Extracts major image contours using gradient operators
Image Sharpening	Enhances image details using a 3×3 convolution kernel
Salt-and-Pepper Noise	Adds random black and white noise to simulate corruption
Median Filtering	Reduces salt-and-pepper noise using a 3×3 median filter
Modern GUI	Dark photo-editor style interface built with EasyX
Status Panel	Displays operation feedback directly inside the interface
BMP Export	Saves processed results as BMP image files

---

Application Preview

Modern EasyX interface with image preview, operation panel, and status display.

---

Processing Results

Original BMP	Grayscale Conversion
Sobel Edge Detection	Image Sharpening
Salt-and-Pepper Noise	Median Filtering

---

Processing Pipeline

24-bit BMP Input
      │
      ▼
BMP Validation
      │
      ├── Check BMP signature
      ├── Check bit depth
      ├── Check compression type
      └── Calculate row stride
      │
      ▼
Load Pixel Data
      │
      ├── Display Original Image
      ├── Convert to Grayscale
      ├── Apply Sobel Edge Detection
      ├── Apply Image Sharpening
      ├── Add Salt-and-Pepper Noise
      └── Apply Median Filtering
      │
      ▼
Save Processed BMP Outputs

---

System Design

ImageFileProcessing
│
├── GUI Layer
│   ├── Draw application window
│   ├── Draw image preview area
│   ├── Draw control panel
│   ├── Draw operation buttons
│   └── Display status messages
│
├── BMP File Layer
│   ├── Read BMP header
│   ├── Validate BMP format
│   ├── Load pixel data
│   └── Save processed BMP files
│
├── Image Processing Layer
│   ├── Grayscale conversion
│   ├── Sobel edge detection
│   ├── Image sharpening
│   ├── Noise generation
│   └── Median filtering
│
└── Display Layer
    ├── Scale image preview
    ├── Render 24-bit color image
    └── Render 8-bit grayscale image

---

Core Functions

Function	Responsibility
calculate_stride()	Calculates BMP row size with 4-byte alignment
read_bmp()	Reads and validates a 24-bit BMP image
save_bmp_8bit()	Saves grayscale results as 8-bit BMP files
save_bmp_24bit()	Saves color results as 24-bit BMP files
convert_24bit_to_gray()	Converts BGR pixels to grayscale
edge_detection()	Performs Sobel edge detection
image_sharpening()	Applies sharpening convolution
addSaltPepperNoise()	Adds random black and white noise
medianFiltering()	Removes noise using median filtering
showImage24Fit()	Displays scaled 24-bit images
showImage8Fit()	Displays scaled 8-bit grayscale images
drawUI()	Renders the graphical user interface

---

Algorithm Highlights

BMP Row Alignment

BMP rows are aligned to 4-byte boundaries. The program calculates the real row size using:

int calculate_stride(int width, int bit_count)
{
    return ((width * bit_count + 31) / 32) * 4;
}

This prevents distorted output caused by ignoring BMP padding bytes.

---

BGR Color Order

BMP stores color pixels in BGR order, not RGB order.

Pixel memory order:
Blue → Green → Red

When displaying pixels in EasyX, the program converts the values correctly:

RGB(R, G, B)

---

Grayscale Conversion

The grayscale value is computed using the standard weighted formula:

Gray = 0.299R + 0.587G + 0.114B

This produces visually balanced grayscale output because the human eye is more sensitive to green intensity.

---

Sobel Edge Detection

Sobel edge detection uses two 3×3 kernels to calculate horizontal and vertical gradients.

int sobel_x[3][3] = {
    {-1, 0, 1},
    {-2, 0, 2},
    {-1, 0, 1}
};

int sobel_y[3][3] = {
    {-1, -2, -1},
    { 0,  0,  0},
    { 1,  2,  1}
};

Gradient magnitude:

G = sqrt(Gx² + Gy²)

Pixels with gradient magnitude above the threshold are marked as edges.

---

Image Sharpening

The sharpening operation uses a 3×3 convolution kernel:

int kernel[3][3] = {
    {-1, -1, -1},
    {-1,  9, -1},
    {-1, -1, -1}
};

This strengthens the center pixel and subtracts neighboring values, increasing local contrast and improving image detail.

---

Median Filtering

Median filtering is used to reduce salt-and-pepper noise.

1. Select a 3×3 neighborhood.
2. Collect the 9 pixel values.
3. Sort the values.
4. Replace the center pixel with the median value.

For color images, the B, G, and R channels are processed separately.

---

Build and Run

Requirements

• Windows
• Visual Studio
• EasyX graphics library
• 24-bit uncompressed BMP image

Build Steps

1. Install Visual Studio.
2. Install and configure EasyX.
3. Open the project source files in Visual Studio.
4. Make sure the test BMP image exists.
5. Build the project.
6. Run the executable.
7. Use the GUI buttons to process the image.

During testing, the input image path was:

C:\CP3\test.bmp

---

Generated Outputs

File	Description
gray.bmp	Grayscale output
edge.bmp	Sobel edge detection output
sharpen.bmp	Sharpened output
noise.bmp	Salt-and-pepper noise output
median.bmp	Median filtering output

---

Testing

Test Case	Operation	Expected Result	Status
TC-01	Read BMP	Original image displayed correctly	Passed
TC-02	Gray Image	Grayscale image generated and saved	Passed
TC-03	Edge Detect	Edge contours extracted and saved	Passed
TC-04	Sharpen	Image details enhanced and saved	Passed
TC-05	Add Noise	Salt-and-pepper noise generated and saved	Passed
TC-06	Median Filter	Noise reduced and result saved	Passed

---

Technical Challenges

Challenge	Cause	Solution
BMP file failed to open	Incorrect file path	Used a verified absolute path during testing
Invalid BMP format	File extension was changed without real conversion	Converted the image properly to BMP format
Wrong color display	BMP stores pixels as BGR	Converted display order to RGB
Row distortion	BMP rows require 4-byte alignment	Implemented stride calculation
Image exceeded window size	Original image was larger than preview area	Implemented image preview scaling
Basic initial interface	Early version lacked visual polish	Redesigned the GUI into a dark photo-editor layout

---

Limitations

• Supports only 24-bit uncompressed BMP images.
• Input path is fixed in the current version.
• Algorithm parameters are not adjustable from the GUI.
• JPG and PNG formats are not supported.
• No side-by-side comparison mode yet.

---

Future Improvements

• Add file selection dialog.
• Add adjustable Sobel threshold.
• Add adjustable noise ratio.
• Add support for JPG and PNG images.
• Add before/after comparison view.
• Add brightness and contrast adjustment.
• Improve performance for very large images.
• Add undo/reset functionality.

---

Academic Report

The complete academic report is included in this repository:

ImageFileProcessing_Report.md

It contains full documentation of the project objectives, system design, algorithms, implementation details, testing results, problems solved, and reflections.

---

Author

HANAN OSSAMA Artificial Intelligence Student Harbin Institute of Technology Student ID: 2025130323

---

Notes

This project was developed as part of the Computing and Intelligent Programming comprehensive practice project. It is intended for academic learning, C/C++ programming practice, and image-processing system implementation.