Published Paper Details:

The paper presents a prototype for a sign language interpreter which can communicate with American Sign Language (ASL). Sign language serves as a visual stimulation of speech and language development. It helps to reduce negative social behaviors, increase social interactions, and develop cognitive structures. With early introduction, sign language provides emotional, social, and academic support for language development. The reliable paradigm is based on building a user-hand gesture-only human-computer interface (HCI). The theories of hand segmentation and the hand detection system are used to construct hand gesture recognition using Python and OpenCV. From the database, the training images are labelled with different names. Each of the images is converted to a binary feature vector. The feature extraction procedure is carried out from the picture collected from real-time video of hand signs using camera and Convolutional Neural Network (CNN). These features are compared with the features of database photos and the system produces output based on the predicted highest similarity. Without a fixed or monochromatic background, this model offers satisfactory accuracy and can be used with the help of an application. The camera is used to record video of the hand gestures of the person, which is then analyzed and converted into textual description. The output is converted to speech, which is heard by the user. The proposed system includes sign language recognition of numbers, alphabets and some common words. Image processing algorithms along with neural networks are used to map the gesture to appropriate text in the training data and hence raw images/videos are converted into respective text that can be read and understood. Such a protocol will undoubtedly be very helpful for bridging the communication gap between those with speaking and hearing abilities and those without them, especially when combined with a large source database.