![]() ![]() In the text detection stage, three text detection models, EAST, DB and PAN, were investigated to select the best text detection model. ![]() In the direction signboard detection, the YOLOv5s model was applied and achieved a precision of 97%and a recall of 96%. This paper proposed a three-stage approach for real-time text extraction: (i) traffic direction signboard detection (ii) text detection and (iii) text recognition. A real-time text extraction from road direction signs is investigated, using input video images acquired from a dashboard camera while driving on Malaysian roads. The road direction sign, which serves as a navigation tool, aids the driver in travelling on the road. Fourth phase is about classification with this the pattern of text is designed and segregated with relevant content and then the last phase is about the postprocessing this tries to clean the record in a specific sense. The overview of working model is it starts with two functionalities such that at first work the Speech to text of the Video is done using Moviepy module and Speech recognition module, now the second work is about Video OCR in this workflow it contains of six phases coming through at initial the video indexing is done and an array of images is developed from the video and then sent to pre-processing here the elimination of noise images and it enhances the image quality and make it ready to next phases now the second phases are segmentation phase in this the image is divided by each line of text into image and then the line is divided to each word into image and then each letter into image from word then the letters were cropped and then the text extracted and next at third step the normalization is performed in this the data is been cleansed and removing of unwanted information is done. It is hoped that this system will optimize work in the pharmaceutical supply chain industry and contribute to the national industry. The system has achieved a 98.04% accuracy rate for OCR and a 100% accuracy rate for MAL numbers and a 90% accuracy rate for product names using Regex. Named Entity Recognition (NER) is also implemented to identify important information from the OCR process. Additionally, too much information for the computer to accurately retrieve from the images exists. The challenges include variations in lighting, image rotation, and different fonts used on the products. Object Character Recognition (OCR) is a commonly used method to extract text characters from images, and in this paper, it is applied to label pharmaceutical products. To address this, a system is needed to automatically retrieve and store information from the image. ![]() Dependable manual labor is necessary to review the product information. A simpler approach would be to scan the image or document and save it as an image file, but analyzing this information can be challenging. The current method of storing pharmaceutical product information involves manual data entry into the computer system. In today's modern age, there is a high demand for storing image and document data to a computer drive for various purposes, particularly in the pharmaceutical industry. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |