Publication:
Edge detection in angiogram image using gradient based method

Loading...
Thumbnail Image
Date
2021-07-01
Authors
Mohamad Wazir, Aida Rohayu
Journal Title
Journal ISSN
Volume Title
Publisher
Research Projects
Organizational Units
Journal Issue
Abstract
Edge detection is a fundamental step in image processing, image analysis, image pattern recognition, and computer vision techniques. It is a process to identify the points in a digital image in terms of pixels where their gradient intensity changes abruptly. Edge detection is a common process in the treatment of medical images. Medical images such as X-ray angiography is a test that uses dye to demonstrate the arteries. Normal arteries are very thin thus invisible to x-ray so a dye which also is known as a contrast agent is injected into the patient’s blood vessels to produce a clearer blood vessels x-ray image. However, the produced images are often low quality due to noises making it difficult for doctors to make accurate diagnoses and treatment decisions. Therefore, the processing of angiograms is necessary to improve the visibility of the image for accurate diagnoses in a short time. In this paper, the angiogram images will be denoised by using the Convolutional Neural Network method. Then, the denoised image will be enhanced using histogram equalization for blood vessels contrast. After that, classical edge operators including Prewitt, Sobel, and Canny are implemented. The edge quality produced will be evaluated by calculating the MSE, RMSE, and PSNR values. The results obtained shows that the Prewitt operator produces better edge quality compared with Sobel and Canny operators due to its low MSE value, low RMSE value, and high PSNR value.
Description
Keywords
Citation