Race Code - PerformProcessingStage usage

These are the various stages of the PerformProcessingStage function available to Race.py.
Each stage is followed by an example of the output.

You can save images using the command SaveImage(image, title).
For example:

SaveImage(someImage, 'saved-photo')

will make a file called saved-photo.jpg in the /home/pi/formulapi/logs directory.
Be aware there are limits on how much data we will upload to the FTP after a race, it would be a bad idea to save lots of images each lap!

Grabbing the camera image (stage 0)

image = GetLatestImage()

Stage 1 - Get the minimums and maximums from an image

rMin, rMax, gMin, gMax, bMin, bMax = PerformProcessingStage(1, image)

Stage 2 - Crop the image

image = PerformProcessingStage(2, image)

Stage 3 - Auto brightness

image = PerformProcessingStage(3, image)

Stage 4 - Find black areas

black = PerformProcessingStage(4, image)

Stage 5 - Erode a channel to remove noise

black = PerformProcessingStage(5, black)

Stage 6 - Split the colour channels

red, green, blue = PerformProcessingStage(6, image)

Stage 7 - Auto level a channel

red = PerformProcessingStage(7, (red, rMin, rMax, Settings.redGain))

Stage 8 - Get the maximum channel for each pixel

maxImage = PerformProcessingStage(8, (red, green, blue))

Stage 9 - Exclude unwanted parts of each channel

red = PerformProcessingStage(9, (red, maxImage, black))

This would usually be followed by an erosion (stage 5)

red = PerformProcessingStage(5, red)

Stage 10 - Merge channels to a single image to make them easy to see

viewImage = PerformProcessingStage(10, (red, green, blue, black))

This is usually for viewing or logging benefit.
Typically image processing / analysis would work with the stage 9 images.

The full listing

Here is the whole sequence laid out in full, saving the results of each stage:

# Grab the image
image = GetLatestImage()
SaveImage(image, '00-raw')

# Get range levels for later
rMin, rMax, gMin, gMax, bMin, bMax = PerformProcessingStage(1, image)
print rMin, rMax, gMin, gMax, bMin, bMax

# Crop the image
image = PerformProcessingStage(2, image)
SaveImage(image, '02-cropped')

# Auto brightness
image = PerformProcessingStage(3, image)
SaveImage(image, '03-auto-bright')

# Find black areas
black = PerformProcessingStage(4, image)
SaveImage(black, '04-black-raw')

# Erode the black areas
black = PerformProcessingStage(5, black)
SaveImage(black, '05-black')

# Split the colours
red, green, blue = PerformProcessingStage(6, image)
SaveImage(red, '06-red-raw')
SaveImage(green, '06-green-raw')
SaveImage(blue, '06-blue-raw')

# Auto level channels
red = PerformProcessingStage(7, (red, rMin, rMax, Settings.redGain))
SaveImage(red, '07-red-leveled')
green = PerformProcessingStage(7, (green, gMin, gMax, Settings.greenGain))
SaveImage(green, '07-green-leveled')
blue = PerformProcessingStage(7, (blue, bMin, bMax, Settings.blueGain))
SaveImage(blue, '07-blue-leveled')

# Get the maximums
maxImage = PerformProcessingStage(8, (red, green, blue))
SaveImage(maxImage, '08-max')

# Exclude unwanted parts of each channel
red = PerformProcessingStage(9, (red, maxImage, black))
SaveImage(red, '09a-red-section')
green = PerformProcessingStage(9, (green, maxImage, black))
SaveImage(green, '09a-green-section')
blue = PerformProcessingStage(9, (blue, maxImage, black))
SaveImage(blue, '09a-blue-section')

# Erode the channels
red = PerformProcessingStage(5, red)
SaveImage(red, '09b-red')
green = PerformProcessingStage(5, green)
SaveImage(green, '09b-green')
blue = PerformProcessingStage(5, blue)
SaveImage(blue, '09b-blue')

# Merge channels to a single image to make them easy to see
viewImage = PerformProcessingStage(10, (red, green, blue, black))
SaveImage(viewImage, '10-final-view')