Shock wave detection on poor-quality shadow image using Image Processing Toolbox

Hi! I am trying to detect shock wave on image using Canny edge detection and other algorithms, but results are not so well.

Shock wave image is dark strip following the light one. The x-position which i want to detect is the boundary between ligt and dark strips.

So if image has good quality I can get it using Canny edge detection algorythm.

But if stripes are barely visible (3 strips, indicated by red arrows), it can't resolve them.

I tryed different image filters (imadjust, filter2, imsharpen) and binarization (imbinarize) with filtering (bwareafilt) but it didn't work.

Here is my code:

filename = 'goodQuality';
extension = '.png';

threshold = 0.4
sigma = 0.1

I = imread([filename extension]);
if size(I, 3) == 3
    I = rgb2gray(imread([filename extension]));
end

[imageHeight, imageWidth] = size(I);
[gx gy]= imgradientxy(I);

cannyEdgeImage = edge(I,'Canny', threshold, sigma);

montage({I, cannyEdgeImage, gx}, 'Size', [3 1]);

So how can I improve image quality to detect barely visible vertical strips? Or should I use another edge detection method?

Answer 1

[Shubh Sahu]

It's tough to get the verticle lines you have mentioned because other lines are more prominent. I haven't tried but I hope gradient will help you to achieve what you want:

[Fx fy] = gradient(F);