Thermal image based object detection And classification in indoor environment To aid visually impaired person

Abstract

In current days, the visually challenged people are accomplishing a great job in our society by competing with the normally sighted people. In spite of various hindrances in their day-to-day life, they overcome those hurdles courageously with the aid of recent technologies. The development of science and engineering has resulted in advent of countless gadgets and software, which plays a significant role in Visually Impaired (VI) person s life. Visually challenged persons experience many issues in terms of mobility, decision-making, object identification etc. To address these issues the assistive aids (white canes, smart canes, ultra cane etc.) and sensors (ultrasonic, depth, vibratory motors etc.) are combined to classify the object or obstacle ahead of VI person. These integrated assistive aids support VI person to function independently and motivate them to face the society confidently. Relying on the requirement of VI person, the objects are identified either in indoor or outdoor environment. This proposed research work focuses on thermal image object detection and classification at indoor room set up. The objects are recognized from thermal images at both day and night time that ensures VI person s navigation even in dark environment. newlineUsually VI person navigates by calculating steps in familiar indoor environment where there is no assurance for their safe navigation constantly. Sometimes VI person or their guests can misplace the objects carelessly, so their safety is not guaranteed even at known environment newline