how to feed the vision assistant a grayscale u8 image?

[123321123]

I'm actually working with kinect camera

I have difficulty with transforming depth array U16 into an image cluster(purple)

when the depth array goes into vision assistant it always get error back(1074396120)(really frustrated T.T)

[Klemen]

So you are acquiring the 11-bit depth image from Kinect and trying to do some image processing on it? What kind of processing are you performing on the depth image?

Try casting the image to 8-bit first and see if that works without errors. You could of course poste more code which would be easier to debug, rather than guessing. But the error -1074396120 refers to "not an image" error.

Best regards,

K

[123321123]

do you have kinect and kinesthesia?

if you have I can post all the vi

[Klemen]

Hello,

i have attached a simple example of Vision Assistant image processing on 8-bit images. It thresholds the 8-bit input image and uses dilation to expand the thresholded objects. The processed image is then displayed.

Best regards,

K

[123321123]

how did you make the purple one u8 input?

[Klemen]

Best regards,

K

[123321123]

let's change an other way 

this time it shows no error but the result isn't what I see in assistant(should be red&black , not all black)

[Klemen]

Hello,

you really should explore some examples first. If you are thresholding an image in two bins, the values returned will be equal to 0 (for values under the threshold value) and usually 1 (for values over the threshold value). If oyu are using an 8-bit image with 256 bins, you will certainly not notice the difference between the values 0 and 1. So in order to see the thresholded area for an 8-bit image in your case, you should use:

Alternative is to change the display pallete to binary.

The red representation in NI Vision Assistant is just an interpretation of the thresholded area.

Best regards,

K