ImageWasher User Guide
Introduction & Principles
Digital images often contain hidden metadata, which is text-based information embedded within the file. This metadata can include sensitive details such as the photographer's personal information, unique image identifiers, and even whether the image was AI-generated.
Such information can pose risks when sharing or uploading images online, including personal data exposure, image tracking, and unauthorized use.
However, simply removing metadata is not enough to ensure an image's uniqueness. ImageWasher goes further by transforming the image itself at the pixel level, creating a visually similar but completely unique image in terms of data.
Parameters & Settings
1. 1. Convert Format
Converts the output image format. When enabled, all processed images will be saved in the specified format.
Format
- PNG: Lossless format, supports transparency. Best for images that need to preserve quality.
- JPG: Lossy format, smaller file size. Quality setting available (0.0-1.0).
- WebP: Modern format, good balance between quality and file size. Quality setting available (0.0-1.0).
Quality
Available for JPG and WebP formats. Range: 0.0 to 1.0. Default: 0.8.
• 1.0 = Highest quality (largest file size)
• 0.0 = Lowest quality (smallest file size)
• Generally, web images (JPG/WebP) are typically saved at quality levels around 0.7 to 0.9, with 0.8 being a good balance between quality and file size.
2. 2. Pattern Overlay
Applies a random noise pattern overlay to the image at the pixel level. This is the core feature that ensures image uniqueness by completely altering the underlying data while preserving visual appearance.
Opacity (%)
Controls the intensity of the pattern overlay. Range: 0 to 100%. Default: 1.5%.
• Lower values (0.5-2%): Almost invisible, minimal visual change
• Higher values (5-10%): More noticeable, but still subtle
• Recommended: 1-3% for best balance between invisibility and data alteration
Note: This is the most important feature for ensuring image uniqueness. Even with very low opacity, the pixel-level data is completely transformed, making it impossible to recover the original image.
3. 3. Border
Adds a border around the image. This adds a separate object to the image, further enhancing uniqueness.
Width
Border thickness. Can be specified in pixels (px) or percentage (%).
• Percentage: Relative to the shorter side of the image (default: 1%)
• Pixels: Absolute value in pixels
• Default: 1%
Opacity (%)
Border transparency. Range: 0 to 100%. Default: 100% (fully opaque).
Color
Border color. Can be selected using color picker or entered as hex code (e.g., #000000 for black).
• Default: #000000 (black)
Random Border
When enabled, generates a random border for each image with random width, opacity, and color.
• When enabled, manual width, opacity, and color settings are ignored
• Each processed image will have a unique random border
4. 4. Brightness
Randomly adjusts the brightness of the image. This modifies pixel values to create variation.
Range (%)
Maximum brightness variation range. Range: 0 to 100%. Default: 10%.
• The actual adjustment is randomly chosen between -Range% and +Range%
• Example: Range 10% means brightness can vary from -10% to +10%
• Lower values (5-10%): Subtle changes, barely noticeable
• Higher values (20-50%): More noticeable brightness changes
5. 5. Saturation
Randomly adjusts the color saturation of the image. This modifies color intensity to create variation.
Range (%)
Maximum saturation variation range. Range: 0 to 100%. Default: 10%.
• The actual adjustment is randomly chosen between -Range% and +Range%
• Example: Range 10% means saturation can vary from -10% to +10%
• Lower values (5-10%): Subtle color changes
• Higher values (20-50%): More noticeable color intensity changes
6. 6. Size
Randomly resizes the image within a specified range. This modifies image dimensions, further altering pixel data.
Range (%)
Maximum size variation range. Range: 0 to 100%. Default: 10%.
• The actual adjustment is randomly chosen between -Range% and +Range%
• Example: Range 10% means size can vary from 90% to 110% of original
• Lower values (5-10%): Minimal size change, barely noticeable
• Higher values (20-50%): More noticeable size changes
• Aspect ratio is maintained during resizing
Usage Tips
- •Pattern Overlay is the most important feature for ensuring image uniqueness. Always enable it for maximum privacy protection.
- •For minimal visual change, use Pattern Overlay (1-2%) with Brightness (5-10%) and Saturation (5-10%).
- •For maximum uniqueness, enable all features with moderate settings. This ensures the image data is completely transformed.
- •All random adjustments are applied per image, so each processed image will be unique even with the same settings.
- •Metadata is always removed regardless of other settings.