Vision impairment affects more than 2.2 billion people globally [1], making their lives challenging as it limits their mobility and independence. Various tools and technologies, such as guide dogs, white canes, and the Phantom Haptic Interface, assist visually impaired individuals. However, these solutions have limitations, such as a restricted range, a ground-limited view, and an inability to identify obstacles or provide information about distant ones. Recently, significant advancements in technology have created new opportunities to address these challenges. Obstacle detection and recognition play a crucial role in developing assistive models. Many existing models lack essential features such as obstacle distance estimation based on proximity to the user, multilingual feedback, and other critical functionalities, making them less efficient and accessible. This research aims to develop a model that recognizes obstacles by processing real-time input from a webcam and estimating their distance from the user. A genetic algorithm is employed to enhance the performance of the obstacle recognition module. The model provides multilingual audio feedback, ensuring its effectiveness across diverse regions worldwide. The proposed model successfully meets all research objectives and demonstrates real-time obstacle recognition and distance estimation to help visually impaired individuals. This innovative solution emphasizes accessibility and safety, helping users navigate safely in their respective environments. The results shows that Navhelp achieved an average distance estimation accuracy of 91.39% and obstacle detection module also demonstrated reliable performance by correctly identifying all objects.