A Study on Improvement of Traffic Sign Detection Accuracy for Blurred Images

In recent years, the demand for image recognition has been increasing. For example, face recognition can be used to unlock lock screens in smartphones, and it can also be combined with a temperature sensor to detect body temperature. It can also be applied to automatic driving, where it can recognize people, traffic lights, and signs using image recognition to increase safety. In automatic driving, images captured by on-board cameras may become blurred due to the shaking of cars. To ensure safety, it is necessary to be able to detect objects stably even when those during driving are blurred. This paper attempts to achieve stable detection.