How can LV determine a mean value of 10 pictures?

[zäpfle]

Hello together,

I would like determine a mean value picture. Because the signal-to-noise ratio is to high. Therfore I get a picture (see Attachment "Original_Picture.png") 10 times (For Loop) and want to calculate the 10 pictures to a 1 mean value picture. I tried it with the Image processing Operator Imaq MulDiv (see Attachment "Example.png"). Unfortunately the 1 mean value picture (see Attachment "Result_Image.png") isn't that what I need.

I hope anybody can help me.

Thanks

zäpfle

[Andrey_Dmitriev]

Simple add 10 images together, then divide by 10, something like that:

Be careful with overflow (therefore I using SGL type for accumulation). If your camera deliver U8, then you can use U16 for accumulation up to 256 images.

(my web cam deliver RGB, so I have extracted Intencity component before accumulation)

Andrey.

[zäpfle]

Hi Andrey,

Thanks for the answer. I tried like your VI (see attachment), but unfortunately it doesn't works. I think the images will added but not divided. Do you see my mistake?

greetings

zäpfle

[Andrey_Dmitriev · Accepted solution]

---

zäpfle wrote:

Hi Andrey,

Thanks for the answer. I tried like your VI (see attachment), but unfortunately it doesn't works. I think the images will added but not divided. Do you see my mistake?

greetings

zäpfle

---

It seems to be that you have overflow. If your images are U8 (8 bit grayscale), then accumulation buffer should be at least I16 (or SGL).

Another point - you may have race condition caused by IMAQ Dispose at the end. Take a note that images theoretically can be disposed before displaying, so you will see nothing at the end.

Andrey.

[zäpfle]

Thanks