Published Paper Details:

Recent advances in the fields of image processing and natural language processing are aimed at creating smart systems that will improve people's quality of life. In this paper, an effective method for text detection and extraction from photographs, as well as text to audio conversion, is proposed. Gray scale conversion is used to improve the incoming image first. Following that, the improved image's text sections are discovered using the Maximally Stable External Regions (MSER) feature detector. To efficiently gather and filter text sections in a picture, use geometric filtering in conjunction with the stroke width transform (SWT). Geometric attributes and the stroke width transform are used to remove non-text MSERs. Individual letters/alphabets are then combined together to identify text sequences, which are ultimately fractured into words. Finally, the words are digitised using Optical Character Recognition (OCR). In the final stage, we convert the identified text to speech using our text-to-speech synthesiser (TTS). The suggested algorithm is put to the test on a variety of images, ranging from documents to nature settings. Promising findings have been presented, demonstrating the suggested framework's accuracy and robustness and encouraging its use..