AbstractThis study focuses on offline script identification of handwritten Gujarati script documents using Optical Character Recognition (OCR) techniques. The goal is to develop an efficient system capable of accurately identifying the Gujarati script from handwritten documents. The process begins with the collection of a diverse dataset of offline handwritten Gujarati script documents. The dataset includes various handwriting styles to ensure the model's adaptability. Ground truth labels are annotated for training and evaluation purposes. Preprocessing techniques are employed to enhance the image quality of the handwritten documents. These techniques involve noise removal, image resizing, and normalization, resulting in clearer and standardized input for the subsequent steps. OCR techniques are then applied to perform the script identification task. These techniques involve the extraction of features and patterns specific to the Gujarati script from the pre-processed images. Machine learning algorithms, such as Support Vector Machines (SVM) or Convolutional Neural Networks (CNN), are trained on the extracted features to learn the script identification patterns. The trained model is evaluated using standard performance metrics, including accuracy, precision, recall, and F1 score. The dataset is divided into training and testing sets to assess the model's effectiveness in identifying the Gujarati script. Once the model is trained and evaluated, it can be deployed for practical use. Given an input handwritten document, the OCR system utilizes its learned patterns to accurately identify and classify the Gujarati script. Overall, this study presents a concise approach to offline script identification of handwritten Gujarati script documents using OCR techniques. The proposed system shows promise in accurately reorganizing the Gujarati script, paving the way for further advancements in this field.