Abstract :
Fine grained classification using recognized scene text in natural images. In this we extract the text from the image and the extracted text is translated to user known language by using language translator. We apply this method in military services. In this the users create their account by giving their details. Now, the user have their user name and password for their further process. The user sends the image to the end user in encrypted type and they can send document also. Encryption is performed by using RSA algorithm. Now, the end user receive the image and they view the image in decrypted type. The end user extract the text from image. The extraction is performed by using OCR algorithm. We subtract the background by background filtering. Once text regions are detected, it perform text recognition. We used two methods for extraction i.e., character extractor and line extractor. The character extractor generates the bounding boxes of words. Each character is compared with ASCII code for translation. In line extractor, it extracts line by line in the image. The extracted text is translated to user known language by using language translator. The accuracy obtained was 85 to 90 percent. N. Bakya | M. Gayathri | R. Krithika "Text investigation for fine-grained object using Context technique" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-2 | Issue-3 , April 2018, URL: https://www.ijtsrd.com/papers/ijtsrd11029.pdf Paper URL: http://www.ijtsrd.com/computer-science/other/11029/text-investigation-for-fine-grained-object-using-context-technique/n-bakya
Keyword :
fine-grained classification, text detection, text recognition, text saliency, language translation