Processing photographic images is important in many applications, among them the development of automated driver assistance systems (ADAS) and autonomous vehicles. Many techniques are used for processing images, including neural networks, other types of machine learning, and edge detection. One common issue with processing these photos is the presence of noise, whether caused by the camera itself or by physical conditions (e.g., weather conditions or dirt on road signs). In this paper, a neural network is used for noise reduction to improve edge detection results and tested with two kinds of noise, Gaussian and salt & pepper noise, and three different edge detection algorithms, Canny, Sobel, and Zhang. Results showed that the noise reduction process was effective in improving performance of the edge detection process, with the exception of conditions where the noise was originally very minimal.