Morning dew, fog, rain, high humidity, or frost – these are just some conditions that can lead to blurry images from your cameras. But don’t worry, you can fix it! 😊
Did you know there are simple methods to find blurry images and improve them? No magical AI required!
Here’s how:
Compute the Laplacian of the image and take its variance. This gives you a sharpness score – the smaller the score, the blurrier the image.
Once you know your data, you can set a threshold to detect blurry images and decide when they need sharpening.
Now, for sharpening methods, there are plenty of options! Let me highlight two:
1️⃣ Unsharp Mask – Yes, it sounds ironic, but this method makes your photos sharper!
2️⃣ Histogram-Based Sharpening – For this demo, I used CLAHE (Contrast Limited Adaptive Histogram Equalization). CLAHE is not only great for blurry images but has many other wildlife applications – I’ll cover these in my next post!
✨ Here’s the simple procedure I followed for this demo:
-
Crop out the white bottom bar.
-
Compute the sharpness of the image.
-
If sharpness < 5.0, run CLAHE. (Thresholds will vary depending on your images and camera setup.)