First of all, a warm welcome to NEW YEAR to all the forum members. As a part of my project work, I should find the area of some objects in an Image. I have done this to a grayscale image in labVIEW recently but unfortunately the results were not helpful because of all object identification is difficult in grayscale image. I would like to continue this on a color image. Would it be possible in labVIEW?
As far I see there are no VI's for color segmentation in labVIEW but I might have wrong though!
Could any one help me.
The standard way to find things in LabVIEW is to use the IMAQ toolkit. However, since I suspect this is a class project, that would be cheating 😉
You can decompose a color into RGB using the Color to RGB VI found in the Graphics and Sound » Picture Functions palette. You will need to process the image yourself to create the 2D array for each color plane, but this is straightforward. Be careful if your images use color maps.
Thank you for your reply. I am currently using Vision options for separating the objects in my color image. Thus, I have started working on "color segmentation VI", in which, the input Color classifier terminal does not make any sense to me. I have gone through some example VI's that came along with Vision but still not clear for me. In the given NI Vision examples, Color classifier(.clf extension file) is directly taken from a path.
what is this color classifier as I don't know and I am new to image processing. How could I get one for my project?
hi i too doing part of my project in image segmentation on grayscale image. since the reults were not upto the mark i wish to calculate the PSNR value to it but i could get a proper result. so help me in calculating the value of it.
If you are looking to separate purely on type of colour Red/Green/Blue, try using the Hue part of the image. If you extract that part of the colour image you will see vastly different results. This is then a greyscale image that you can threshold just as you would for intensity.
If it is a specific colour then use the ColorThreshold VI to do this.
Generally working with RGB values for a colour image is not good. Try using the HSV colour space and it is much easier to identify regions of colour. have a look at the Color Spaces section of the NI Vision Concepts Manual.