Despite the intense research efforts in character recognition field, ability of machines to replicate human reading capabilities is still underneath. While OCR systems for Latin scripts have achieved high accuracy, OCRs for Arabic script-based languages, such as Urdu and Sindhi, still face significant challenges, particularly due to complex cursive writing styles like Nastalique. Sindhi, a language rich in literature and poetry, uses a script similar to Arabic but includes additional characters. This makes available OCR systems inadequate for accurate recognition of Sindhi. Among the various stages of OCR development for these scripts, the segmentation stage is especially critical, as accurate segmentation is essential for the success of subsequent processes like feature extraction and classification. This paper introduces a technique for segmenting printed Sindhi ligatures, a crucial step in Arabic script-based OCR systems. The process of ligature segmentation in recognition systems faces several challenges, including merged ligatures, horizontally overlapping and diacritic association. To address these challenges, we have applied specific algorithms and techniques to images of printed Sindhi text.