Table 1 Tabulated summary of all pre-processing technique
Preprocessing technique | Description | Application in Stone Inscriptions |
|---|---|---|
Thresholding | Techniques like Otsu’s Method, Niblack, and Sauvola adaptively establish thresholds, aiding in segmenting text from backgrounds and handling varied lighting conditions | Essential for segmenting inscribed content and adapting to diverse lighting conditions |
Median filtering | Replaces pixel values with the median of neighboring pixels, reducing sudden noise caused by imperfections on stone surfaces | Effective in removing unwanted noise from scratches or imperfections on stone surfaces |
Gaussian blur | Smooths images without compromising the integrity of inscribed text or details, reducing minor irregularities and noise | Useful for refining stone inscription images without losing essential details |
CLAHE and gamma correction | Techniques optimizing contrast, controlling noise, and enhancing image quality, specifically adapted for stone inscriptions | Enhance readability and preservation of historical content in stone inscriptions |
Resizing | Adjusts image dimensions for efficient computational processing, optimizing memory usage and retaining essential content details | Manages large image sizes while retaining critical information for analysis |
Grayscale conversion | Converts images to grayscale, emphasizing contrast and intensity variations, focusing on the legibility of inscribed content | Helps in emphasizing contrast and clarity, aiding in the interpretation of inscriptions |
Bilateral filtering | Smooths homogeneous regions while preserving edges and details within inscriptions, enhancing overall image quality | Improves image clarity while maintaining the integrity of intricate details |
Morphological operations | Erosion, dilation, opening, and closing refine image structures, aid in edge detection, and manipulate inscription features | Vital in shaping, refining, and enhancing structural details within stone inscriptions |
Edge detection | Techniques like Canny, LoG, and Sobel detect and highlight edges, aiding in capturing intricate details within inscriptions | Essential in uncovering and analyzing crucial features and details within stone inscriptions |
K-means clustering | Segments inscribed content or distinguishes various elements within images, aiding in identifying patterns or text from non-text regions | Helps in segmenting and identifying different components or patterns within stone inscriptions |
Denoising (Fast means, NLM) | Techniques reducing noise and enhancing image clarity, crucial for preserving critical details within the inscriptions | Mitigates noise interference, allowing for clearer analysis and interpretation of inscriptions |