Advancements in computer vision-based assistive technologies for visually impaired

The advancement of assistive technologies for visually impaired individuals has seen significant strides, leveraging the power of computer vision-based object recognition and detection. This literature review explores various approaches aimed at enhancing navigation, obstacle detection, and overall mobility for the visually impaired. By utilizing methods such as real-time object detection, deep learning algorithms, and wearable devices, these technologies convert visual information into comprehensible formats through image-to-text and text-to-speech processes. Recent research efforts have focused on integrating these technologies to provide real- time auditory feedback, thereby enhancing spatial awareness and safety. For instance, mobile applications like "Smart_Eye" and wearable devices based on platforms such as Raspberry Pi employ advanced sensors and machine learning to identify and describ e objects, providing crucial auditory cues to users. Systematic reviews of wearable electronic aids and technologically advanced guidance systems underline the importance of consistent detection accuracy and user adaptability. Despite the promising developments, challenges such as spatial biases, computational intensity, and user interface complexities remain. This review consolidates the findings from various studies, presenting methods used, limitations encountered, and the potential impact of computer vision-based recognition systems to assist visually impaired users. By converting image data to textual descriptions and subsequently into speech, these technologies offer comprehensive support, significantly improving the independence and qualit y of life for visually impaired individuals. Overall, the integration of computer vision, deep learning, and real-time feedback systems marks a transformative phase in assistive technology, providing visually impaired individuals with enhanced tools to navigate and interact with their environments effectively.