I have a 3D CT image of a car engine in raw form with no header information. When loaded into numpy array as 16 bit unsigned integer, I noticed that the values range between 0 to 52000. Is this normal for a CT Image? When viewing the image, I notice a lot of cloud like noise around the parts. I'm trying to extract features using a deep learning method. This is my first time working with CT Images. What pre processing is neccesary for such CT Images?
As 16-bit images contain values from 0 to 65535, your data is probably just scaled to fit that range. The cloud like noise could be beam hardening, especially since you work with a heavily absorbing sample. Reducing beam hardening is best done during the image acquisition or if not possible in the 3D reconstruction. If you only have access to the 3D images there are probably no simple solutions.
Ilian Häggmark Thank you for the reply. Thats a great insight. Also, yes, I only have access to the 3D images. Is thresholding acceptable? since the cloud like noise is usually of low intensity, could I maybe remove it by thresholding?
With the existing 3D images, the feature extraction method I use produces mostly noise. When I try the same with medical Images, the features extracted are great. Any good approach to tackle this?
It's typically hard to remove beam hardening artifacts with such a simple approach as thresholding, but if it can help the feature extraction I would try it.
As for the deep learning method. This is not my area of expertise, but I would start by thinking about the following: how was your method trained (on what data) and how does the SNR compare?
- Badges
- Science topic
- Similar topics
- Computer Science
- Images