• Yashwanth G
• Karthik R
• Anusha J
• Yakshitha K C
• Yashaswini N L

Image processing is used in various industries and is one of the most advanced technologies used in Google, the medical field etc. google vision is one of the most used APIs in our daily lives for image labelling, face and landmark detection, optical character recognition (OCR), and tagging of explicit content, etc. The widely used google photos and iCloud use image processing to recognize the scenario of the photo and facial recognition. Image processing can help the visually impaired describe their surroundings. A voice-based image caption generation which is built using encoder-decoder architecture is used to describe the image. Images of the surroundings are used to generate captions which will be read aloud to the visually impaired so that they'll get a stronger sense of what's happening around them. As an encoder, we have a pre-trained Resnet50 model where ResNet-50 is a convolutional neural network which is 50 layers deep. This network learns rich feature representations for a large range of images. Besides extracting high-level features from images using Resnet50, we also maintain the image colour composition using OpenCV techniques which also helps the model to extract the features from small components within the image. As a decoder, we have an LSTM network. Long remembering (LSTM) is a recurrent neural network (RNN) architecture and contains the Time Distributed Layer that's, it can process not only single image data points but also entire sequences of image data.

Deep learning, Resnet50, LSTM, Flutter, Encoder-Decoder, Flicker8k

"Automated Image Caption Generator For Visually Impaired", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.9, Issue 4, page no.h496-h501, April-2022, Available :http://www.jetir.org/papers/JETIR2204766.pdf

"Automated Image Caption Generator For Visually Impaired", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.9, Issue 4, page no. pph496-h501, April-2022, Available at : http://www.jetir.org/papers/JETIR2204766.pdf