Python OpenCV: Blurt Detect

Python OpenCV: Blurt Detect

I settled a Raspberry pi for taking pictures for months to recording the pictures of my Eco-Tack. It is fun to see the changing of the plants in the tank after combined the pictures. But the problem is their are few blurt images in those hundreds of pictures. So, I wonder it is really cool that detect the blurt pictures by using python. So, I found two posts from internet and try to find the suitable one for me.
Two reference posts:
技术挖掘者; 2019
Dontla; 2019

So, there have 574 photos in total and I have no idea which one is blurt images.

技术挖掘者

Blurt	Clear

Dontla

Blurt	Clear

Compare

By comparing the result of those to codes, I found that they used the exact same way: cv2.Laplacian to calculate…
I should really check the raw codes before executing them.

First	Second

The simplified codes should be:

import cv2

imagePath = "img.png"

image = cv2.imread(imagePath)
## 将图片转换为灰度图片
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
## 计算灰度图片的方差
fm =  cv2.Laplacian(gray, cv2.CV_64F).var()
print(fm)

很不实用 = = 真的单独一个算法来这么检测, 太水了.

Plus

So, there are another good post about the blurt-detection by using python Opencv.
It applied ‘Brenner’, ‘Laplacian’, ‘SMD’, ‘SMD2’, etc to calculate. If you are interested in it, go to check FUNNY AI.