Hello,
I wanted to plot a graph of the bubble size distribution from this image.
This image is taken from a slow-motion video (the video was taken with a mobile phone) where the bubbles enter from the bottom to the top.
Various software such as Image J, Fuji, bubble analyzer and Motic Imageplus 3.0 were used to analyze the image. However, I could not accurately identify bubbles with these programs. (Of course, if I have done things correctly with the software)
What program do you recommend for checking this image for all bubble identification?
According to the image, do you think it is possible to detect all the bubbles using the mentioned programs?
Or do you know another way with a mobile phone to take pictures of bubbles moving at high speed?
Thank you in advance for your guidance in this matter.
Hi Fatemeh Mahmoudian;
You need to adjust the shutter speed according to the vertical speed of the bubbles so that all the bubbles appear stationary in your image. Of course, bubbles of different sizes probably have different speeds, but if their sizes are not much different, you will see almost most of them stationary with this method.
I think with MATLAB image processing toolboxes you can analyze the obtained images to determine the size of the bubbles. I don't know, but I think the MATLAB image processing toolbox is a complete thing.
I agree with Seyed Mohammad Mousavi.
To count the number of bubbles within a photo of a liquid using MATLAB, you can utilize various toolboxes and approaches. For example:
- Image Processing Toolbox. You can use segmentation methods to separate the bubbles from the background and then count the number of distinct regions or objects in the image.
- Computer Vision Toolbox offers algorithms and functions specifically designed for computer vision. You can use object detection or segmentation methods to detect the bubbles and then count their number.
- You can utilize "Image Segmentation" techniques. You can employ image segmentation algorithms such as Otsu's method or the Watershed algorithm to separate the bubbles from the rest of the image. Subsequently, you can use object counting or connectivity analysis operations to determine the number of bubbles in the image.
- You can utilize "blob analysis" to identify and count bubbles as connected objects in your image. This can be done using functions such as "bwlabel" to label connected objects and "regionprops" to extract blob properties such as area or perimeter.
You can also use the Python language to write a specific program. There are several specific libraries in Python that can be used to count "closed areas" in an image. Some examples are: "OpenCV" offers many image analysis functions and algorithms that can be used for detecting and counting areas. You can use segmentation functions like cv2.findContours(), along with functions like cv2.minEnclosingCircle(), to detect and count circles in the image. "scikit-image" is a Python library dedicated to image processing. It provides advanced functions for image segmentation, such as skimage.segmentation, which can be used to find and count closed areas in the image. "SimpleCV" is a simplified interface for OpenCV in Python, designed to streamline the image processing workflow. It offers functions for counting, such as image.findCircle() and image.findBlobs(), which can be used to find and count circles and other shapes in the image. "Mahotas" is a Python library for image processing that provides a variety of segmentation and image analysis algorithms. You can use functions like mahotas.label() to label connected regions in the image and then count the number of circles or other closed areas.
Hi Fatemeh Mahmoudian,
My ex-colleague used high speed camera to take pictures and then import them to the software Image Pro-plus to analyze and obtain the the bubble infromation.
Pleace check the link of his work:
Article Anodic Bubble Behavior and Voltage Drop in a Laboratory Tran...
regards,
xiaozhen
It appears that some of the bubbles are large compared with the tank dimensions used. If that is so, the large bubbles will be wall affected and may well be cylindrical rather than spherical. Whatever the case, I would always advocate doing manual measurements of the dimensions before using analytical methods: this ensures that you know whether the distribution is normal or skewed and allows you to explore various mean diameter formulae. Computer assessment is great once you are clear about the basics, but will not resolve wall effects or shape effects unless you program these in. A magnifying glass and a ruler are good first level tools.
- Badges
- Science topic
- Similar topics
- Counseling Psychology
- Confusion