I am using opencv
to detect person in live video feed. I need to save the image of the person detected. But here the person is not standing and is keeps moving due to which when I am about to save the image, it is saved in very blurry format, just like below image:
As you can see the image is not very clear and has a lot of blurriness into it. Face is also not clear. Is there anyway we can remove the blurriness from image. Thanks
We use sharpen() function to sharpen an image.
You use a Gaussian smoothing filter and subtract the smoothed version from the original image (in a weighted way so the values of a constant area remain constant). cv::GaussianBlur(frame, image, cv::Size(0, 0), 3); cv::addWeighted(frame, 1.5, image, -0.5, 0, image);
To smoothen an image with a custom-made kernel we are going to use a function called filter2D() which basically helps us to convolve a custom-made kernel with an image to achieve different image filters like sharpening and blurring and more.
You can try sharpening the image using cv2.filter2D()
and a generic sharpening kernel
Here are other sharpening kernels you can experiment with
import cv2
import numpy as np
image = cv2.imread('1.jpg')
sharpen_kernel = np.array([[-1,-1,-1], [-1,9,-1], [-1,-1,-1]])
sharpen = cv2.filter2D(image, -1, sharpen_kernel)
cv2.imshow('sharpen', sharpen)
cv2.waitKey()
As suggested by @nathancy you can try with morhological based operations, this works in all the cases but it requires filter dimension tuning according to image shape, noise level and it also leads to drop in image quality.
Recently, Generative Adversarial Networks (GAN) has also got attention for regenrating the images and seems to be promising in image quality enhancement.
This article(https://medium.com/machine-learning-world/deblur-photos-using-generic-pix2pix-6f8774f9701e) has described a GAN based (based on pixtopix model) solution for image deblurring, this may work for your case too.
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With