Image comparison from different angles

Hi

I'm currently working on an algorithm to compare images from different angles and brightnesses. I have searched many papers and threads, but I didn't find any suggestions or ideas to solve my problem.

The target of my research is an algorithm that locates and highlights the main differences in the two images.

Here are the two example Images that I'm working with.

Can anyone help me with this?

[Image Analyst]

Please insert your images here with the picture frame icon.

Then define what "compare" means to you. What algorithm have you researched and developed? Or do you simply want to use something that's already out there, and not develop anything new yourself?

[Andre Willems]

1. done.

We use one picture as should condition. The second one should be the is condition.

The pictures are taken with a drone. The variatons of the positioning to look at our is condition cause different angles and brightnesses.

I'm not at the point of actually creating an algorithm. It's more like preliminary study. I'm lookin for methods to create a new one. If there are some functions i can use, that would be great.

Pre-Processing:

• Identification of weather induced noise with a non-linear-regression model
• edge detection to eliminate brightness effects

Acutal Processing

• my Problem

Post-Processing

• Highlight and display the differences

Answer 1

[Image Analyst]

Of course the simplest way is to just subtract the images to find pixels that are different. You can also take the histograms of each, and they will be different for the images you posted. There are of course other ways. Search the image processing literature for relevant keywords: Vision Bibliography

[Andre Willems]

thank you. I will have a look for extra literature