An Improved Image Enhancement of Road-Lanes and Paint-Signs Using Verilog HDL

In this paper, a better enhancement technique for lane and paint sign detection under harsh noisy sunlight conditions are implemented in Verilog HDL. The image enhancement technique proposed can be used in real-time feature extraction for road lanes and paint signs. The design module uses thresholding techniques to highlight pixels that define the lanes and signs. The method of inverting an image is used to generate grayscale images in this design, and is pipelined with the thresholding technique. The failure in attaining maximum efficiency resulted due to reduced noise detection percentage. The noise can be any form; one example can be of sunlight creating a pattern of exposure levels in an image frame. The model designed in this paper has improvements in the detection of these noises and eliminating them to finally have only important pixels required for detection. The model is verified for errors and tested under different conditions to predict the best threshold values to choose, depending on the environment conditions.