Okay, so today I wanted to mess around with a couple of different image processing techniques, specifically looking at edge detection. I’ve heard of both “Kenin” and “Garcia” but never actually tried them, so I figured, why not give them a shot and see what happens?
First, I needed some images to work with. I grabbed a few random pictures from my photo library – nothing special, just some everyday shots with various objects and backgrounds. I made sure to get a couple that was well-lit with some images that has less-than-ideal lighting conditions.I think I got 3 or 4 photos in total.
Kenin Method
I started with the “Kenin” method. Honestly, I had to do a little searching online. I ended up reading a bit of this article here, not to copy, but I think it did help me get a better understanding. It describes the method as something related to find edges and trace boundaries in images and then generate a vector file. But I have no idea what this description means. Maybe I will try to use python and Opencv to do it?
So, my approach might seem a bit all over the place, but bear with me. First I fired up my usual image editing tool and loaded one of the images. I used to use Photoshop but this time, I’m trying out GIMP. After converting my image to greyscale, I looked into the tools and played around, fiddling with various settings.I played around with the levels, trying to get a decent balance of lights and darks.
I messed around with blurring, I did a Gaussian blur, trying a few different radius values. It took some tweaking, but eventually, I got something I was kind of happy with. Then, I started trying different edge detection filters within the tool. It was all a bit of trial and error, to be honest.
Garcia Method
Next up was the “Garcia” method.I remembered Garcia is an edge detection filter, there has something called “gradient magnitude”. Again, I used a combination of searching online. So, I went back to my image editor with a fresh image.
I converted the image to grayscale and then applied what I thought was the “Garcia” method. I found this other filter that seemed similar, focused on gradient. Honestly, I’m not 100% sure if I did it the “correct” way, but hey, it’s all about experimenting, right?
I tried to keep the settings somewhat consistent with what I did for the “Kenin” method, just to see if there was a noticeable difference in the results. I used simliar levels. The image was already pretty sharp, so I did not do more blurring this time.
Comparing the Results
So, after all that, what did I actually get? Well, both methods definitely highlighted edges in the images, but in slightly different ways.
- Kenin: The “Kenin” method created sharper edges, the image looks good.
- Garcia: “Garcia” method, the one using the gradient filter, created more strong edges and the differences are obvious. I like this more.
It was a fun little experiment! I definitely learned a bit more about how these different edge detection techniques work, even if I’m still a bit fuzzy on the exact details. It just goes to show, sometimes the best way to learn is to just jump in and start playing around.
