Dataset Image Augmentation Generator

A powerful desktop application that creates multiple variations of images to expand machine learning datasets. Select a folder, choose from 30 augmentation techniques, and generate new images with one click.

---

✨ Features

• 🚀 30+ Augmentation Techniques – Comprehensive set of transformations
• 🖱️ One-Click Generation – Select options and click start
• 📦 ZIP Download – Automatically packages augmented dataset
• 📊 Progress Tracking – Real-time progress bar and logs
• ✅ Select/Deselect All – Easy option management
• 🖼️ Multiple Formats – Supports JPG, JPEG, PNG

---

🎮 Usage

1. Launch the application (double-click .exe or run the Python script)
2. Select Input Folder – Choose folder containing your original images
3. Choose Output Location – Use default or browse custom folder
4. Select Augmentations – Check/uncheck desired techniques
5. Click START AUGMENTATION
6. Download ZIP – Click download button to save augmented dataset

---

🔧 30 Augmentation Techniques

Geometric Transformations

#	Technique	Description
1	Rotation	-3° to +3° rotation
2	Horizontal Flip	Mirror image horizontally
3	Vertical Flip	Mirror image vertically
4	Random Shear	Shear transformation
5	Random Zoom	Zoom in/out (0.8x to 1.2x)
6	Random Translation	Shift X/Y by -20 to 20px
7	Perspective Transform	3D perspective warp

Noise & Blur

#	Technique	Description
8	Gaussian Noise	Add random Gaussian noise
9	Salt & Pepper Noise	Add salt & pepper noise
10	Motion Blur	Simulate camera motion
11	Gaussian Blur	Smoothing blur

Color Adjustments

#	Technique	Description
12	Brightness Adjustment	Adjust image brightness
13	Contrast Enhancement	Enhance image contrast
14	Color Jitter	Random color variations
15	Gamma Correction	Adjust gamma levels
16	Channel Shuffle	Random RGB channel permutation
17	Posterize	Reduce color palette
18	Random Grayscale	Convert to grayscale

Advanced Techniques

#	Technique	Description
19	Elastic Deformation	Rubber-sheet deformation
20	Cutout	Random black square patches
21	CLAHE	Adaptive histogram equalization
22	Edge Enhancement	Sobel edge detection
23	Histogram Equalization	Global histogram equalization
24	Fourier Noise	Frequency domain noise
25	Random Occlusion	Random colored patches

Deep Learning Methods

#	Technique	Description
26	Mixup	Image blending (placeholder)
27	CutMix	Cut and mix patches (placeholder)
28	CutMix	Advanced mixing technique
29	Fourier Noise	Frequency domain augmentation
30	Posterize	Reduce bit depth

---