This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(86.4, 9%, 18%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EPM%3C/text%3E%3C/svg%3E)
I am working on my final year project using Python and VS Code.
This involves using the live depth data from a Kinect V2 which then gets processed to predict the motion of an object.
The steps the image shall go through include:
As I am using a Microsoft Surface Pro 4 (i7) even the background subtraction step makes my device chug (512*424 16bit images).
I am wondering can I use Azure Functions to offload the processing (e.g. send a frame to azure to compute then return the computed data) or is a Virtual Machine a better option?
Questions:
Can I use Azure Functions or should I use a VM for this?
If Azure Functions can be used:
Can an Azure Functions be passed data, such as an image frame/array?
Can an Azure Functions return data, such as an image frame/array?
How?
I have no real experience with Azure or VMs in general as I am used to running scripts locally.
Any help would be much appreciated, thanks.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(89.60000000000001, 68%, 18%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EJM%3C/text%3E%3C/svg%3E)
Hello @Padraig McGloin , Welcome to Microsoft QnA. Thank you for posting your query here. We will review and update at the earliest.
Hello @Padraig McGloin , you may check https://stackoverflow.com/questions/61790607/how-do-i-receive-multipart-form-data-in-azure-function/63305138#63305138 and see if this helps at all.
@JayC-MSFT thanks for getting back to me.
Exactly what I am trying to do is this:
def create_Mask(self, frameData, backgroundSubtractor): # send frame (kinect depth [16bit], 217088 values) and background subtractor(open cv KNN)
fgmask = backgroundSubtractor.apply(frameData) # apply subtractor to frame
flattenMask = [value for element in fgmask for value in element] # flatten mask
flattenMask = np.array(flattenMask) # transform into numpy array
frameMask = []
for val in np.nditer(flattenMask): #create mask of 1s and 0s
if val == 255:
frameMask.append(1)
else:
frameMask.append(0)
frameMask = np.array(frameMask) # tranform into numpy array
return frameMask # return mask (numpy array [1 or 0], 217088 values)
Taking this function and getting an Azure function to do this functions compute, is it even possible to do something like this using an Azure function?
Hello @Padraig McGloin , you can do that but from the performance point of view I would suggest to consider blob storage ( you may refer to the example I shared).
If this answer helped, please Accept the answer and Up-vote so that it helps others in the community.