#Bytes Array Conversion
spare depot
not sure what your original question was
but heres this
conversion + compression/decompression all there
if you need more compression, you do delta compression, which might have some loss to it
not sure if this is what you're looking for
ivory charm
If you can help out with the conversion / decompression, that'd be great.
i have absolutely no information about the byte array except for the numbers in the array. if you wanna provide the link where you got that array or explain how you got it, then it would be easier to figure out
empty girder
This is essentially the message the object gets every update.
(For the record here, think of messages like a simplified struct / class)
Doing a bit more digging revealed this sample code for CV2 use with the compressed image message.
Bringing this up as reference for how the message is typically used with regards to the data parameter
And a bit more digging revealed the info regarding the formatting for the image sent
imgcomp =
ROS CompressedImage message with properties:
MessageType: 'sensor_msgs/CompressedImage'
Header: [1x1 Header]
Format: 'bgr8; jpeg compressed bgr8'
Data: [30376x1 uint8]
So it's a bgr8 image, essentially.
(Which is why, as shown above with CV2's sample, there was no conversion needed; CV2 has bgr8 by default)
So maybe it's Unity getting confused trying to read the uint8[] array (that's a bgr8 image) as a uint32[] array (rgb32 image)?
Thus why Unity can't output the image out of the box and shows that red question mark.
I did manage to find this post on Stack Overflow... With no responses...
Is there any way to change the color formatting of a texture2d variable?
ivory charm
okay so I just want to make sure I am completely understanding this, you are getting a cv2 image as a byte array and trying to display that as a texture2d in unity, correct?
empty girder
I'm getting a ROS topic which contains the uint8[] data variable, and I'm trying to display it as a texture2d in unity.
(Brought up CV2 just for reference and the possible explanation of the formatting, my project doesn't involve CV2; apologies about the confusion)
ivory charm
where is the original image that is being converted to your byte array coming from? and how is it being converted to the byte array?
empty girder
Basically it's from a live video feed from a drone; and the camera sensor is publishing the data to a topic that sends out the message.
From there I'm using ROS Sharp to take in the published message
ivory charm
so after doing more reading and testing, it looks like it shouldn't need any sort of conversion or anything like that. So this byte array you are using for testing is likely the issue. Are you absolutely certain it is a valid image?
I was able to find this https://github.com/siemens/ros-sharp/blob/master/Unity3D/Assets/RosSharp/Scripts/RosBridgeClient/RosCommuncation/ImageSubscriber.cs and all they are doing is calling the LoadImage extension method then applying the texture
empty girder
Yeah a peer of mine used the same code base
Kind of odd, since the topics remain the same.
Copied the code once more, still getting the same question mark