Object Detection for Signboard

Abstract

Aim of this project is recognize occulated images in signboard and recognize the signboard to help impaired persons and to avoid accident. Now a day many accidents are, occur the main reason for accident is traffic to control the traffic signboard is important one, but occulted signboard lead to traffic and accidents. Physically impaired and invisible persons did not saw that signs so, this system is very useful to their. Then autonomous vehicles is work efficiently using in the system, the signboard is automatically recognized and this proposed system guide the autonomous vehicles. This system recognize the input image and give alertness to the driver. Input image is taken from camera that is placed on in front of the vehicle that capture the image that is consider input image. Then the image is covert a gray scale image then resize the converted image using resizer process, and remove noise using any filter, after this process feature extraction and feature reduction method are used here for remove unwanted facilities and add more wanted feature and facilities. Then DNN process is important role in the system it is used to recognize speech signal and image signal. The main goal of this project is analyze the input image and recognize it and compare that input image with data set that  are stored in data base the data is matched with data set give alert to driver.