<p>Since its debut, optical character recognition (OCR) technology has advanced dramatically, turning various document and image formats into editable and searchable data. This paper offers a thorough examination of the approaches and results concerning OCR systems, concentrating on handwritten, printed, multilingual, online and offline, natural scene, and video Urdu OCR systems. To guarantee the validity and trustworthiness of the results, the research methodology comprised a painstaking search and selection procedure, stringent inclusion and exclusion criteria, quality assessment, and data extraction. Our findings show that there are distinct difficulties and different degrees of accuracy and efficiency for every kind of OCR system. While printed OCR systems exhibit superior accuracy but are impacted by typeface variances and document quality, handwritten OCR systems are significantly challenged by the intricacy of unique handwriting styles. The report also emphasizes the differences between offline and online OCR systems, highlighting the benefits and drawbacks of each in various use scenarios. The evaluation of multilingual OCR systems reveals areas that need improvement while also demonstrating advances in handling a variety of languages, scripts, and dialects. While there have been significant breakthroughs in natural scene OCR systems, these systems still face challenges from environmental elements like background noise and lighting when processing text from pictures and outdoor images. Furthermore, the investigation of video Urdu OCR systems highlights the complex constraints related to real-time text extraction from moving images and the unique difficulties presented by the Urdu script.</p>