Machine Vision

cancel
Showing results for 
Search instead for 
Did you mean: 

how to apply vertical and horizontal projection?

Solved!
Go to solution

Dears,

I hope that anyone can help me to solve this problem...

I need to apply vertical projection on image that contain a vehicle with plate to determine the propably location of the car plate.

please find the attached 2 images 17.jpg is image which i want to apply horizontal projection, v-h[projection].png is what i want to do like it

 

thank...

 

Download All
0 Kudos
Message 1 of 5
(5,824 Views)
Solution
Accepted by topic author GeniusIDS

Looks like IMAQ LinearAverages to me.  If you multiply the averages by the number of rows or columns, you will get the count for a binary image.

 

Bruce

Bruce Ammons
Ammons Engineering
Message 2 of 5
(5,819 Views)

Thanks

That solve one of my question, Sobel edge detection find all edges in the image. i want to seperate the vertical edges from horizontal edges.

 

Thanks ,,,,,

0 Kudos
Message 3 of 5
(5,755 Views)

Hello,

 

if you want to seperate vertical and horizontal approximations of the derivatives, use linear filter (IMAQ Convolution operator), specifying two different convolution kernels - one for vertical and another for horizontal direction.

 

For example, see here:

 

http://forums.ni.com/t5/Machine-Vision/creating-double-precision-image-from-uint8/m-p/2796260#M41746

 

Best regards,

K


https://decibel.ni.com/content/blogs/kl3m3n



"Kudos: Users may give one another Kudos on the forums for posts that they found particularly helpful or insightful."
0 Kudos
Message 4 of 5
(5,746 Views)

Thanks Klemen

 

Im trying to find way to extract car plate from image

i use gaussian -> Sobel -> detect rectangle but accuracy not good

i use gaussian -> match pattern but the pattern not available in all images

i use sobel -> binary -> vertical/ horizontal projection not all edges detected

 

please if  you can share with me some techniques it will be good for me

 

Thanks

Download All
0 Kudos
Message 5 of 5
(5,708 Views)