A quiz on the stomata-counting experiment

Adrian F. Clark

1. What is the main purpose of this experiment?

   • to let you understand that OpenCV can be used to build graphical user interfaces

   • to see how contours centres are found
   • to practise writing calls to cv2.inRange
   • to count the number of stomata

   The main purpose of the experiment is to help you understand how OpenCV can be used to build useful GUIs. Although the main aim of computer vision is to take the human out of the task, it is sometimes (as here) necessary to involve them. A GUI is the currently conventional way to do it.

2. What are stomata?

   • microscopic pores on leaves through which transpiration takes place

   • spores of a fungal disease
   • lumps caused by a virus
   • microscopic wasp larvae

   Stomata are the parts of a leaf through which gases pass, forming an essential part of plant growth. As a rule of thumb, the more stomata and the larger they are, the healthier and quicker-growing the plant is.

3. Which RGB slider has the most effect on the detection of stomata?

   • blue
   • green
   • hue

   • red

   In my experience, setting the red and green sliders to encompass most values allows the blue slider to control the effectiveness of segmentation.

4. Which Python package is used to handle the command line?

   • sys

   • argparse

   • taskfile
   • os

   The standard Python package for handling the parsing of the command line is argparse, and it is available on all platforms. Yes, even Windows.

5. If you are using count-stomata.py on a system without a right mouse button, such as a Macintosh, what does the program let you do instead?

   • hold down the control key

   • hold down the shift or control key

   • hold down the windows key
   • hold down the ALT key

   You are able to hold down either the shift or control keys; the relevant code is in the callback that handles events.

6. Why is the variable 'me' given as an argument to the call that creates a slider?

   • it contains the image name
   • it contains the command line
   • the program name is needed to create a slider

   • there may be more than one window

   There may be more than one window in the graphical user interface, so you have to specify in which window you want the slider to appear. The choice of window name is arbitrary but here it is the program name.

7. What does a space character do in the graphical user interface of count-stomata.py?

   • it controls the segmentation process
   • it quits the program

   • it moves on to the next image

   • it tells the program to count the number of stomata

   The space bar tells count-stomata.py to move on to the next image given on the command line.

8. How are the stomata positions found?

   • the centre of the bounding box of the contour

   • using the ratio of the first two moments
   • the centre of the minimum enclosing circle
   • the centre of the minimum enclosing ellipse

   To be honest, all of these answers could work but the code uses the middle of the bounding box

9. How is the slider position communicated to the segmentation code in count-stomata.py?

   • using a global variable
   • using a local variable

   • using a callback

   • there is no need to do anything

   a callback is defined when the slider is created and this sets a global variable with the value from the slider, and that is passed into cv2.inRange.

10. Which colour space is most effective at segmenting stomata?

    • RGB

    • HLS
    • HSV
    • monochrome

    My experience is that, unusually, RGB space is the most effective at segmenting the stomata.