A Spreadsheet-Based Manipulative to Support Teaching Convolution
Session Format
Presentation Session (45 minutes)
Location
Room 2904 A
Abstract for the conference program
Blurring an image is fairly common but the question is, how does one blur a digital picture? The simple answer is to apply a blur filter…but how does that blur filter blur a picture that is stored as numbers? If it were a physical charcoal drawing the artist would smear the charcoal out with his finger to transform a sharp, dark line into a wider, lighter gray line but your camera produces images that are collections of numbers. How does that work?
Convolution calculations underlie many of the filters found in bitmapped image processing programs like GIMP and Photoshop. In Georgia Gwinnett College’s Digital Media class we “peek under the covers” to see how things work because we believe that knowing a little about what is going on helps dispel the idea that it is all magic performed by wizards. In the process of doing that the hope is to empower the student by demonstrating that he can understand what is going on and is not simply a slave to the gods of computing. One of the “under the cover peeks” we take is to look at these convolution calculations and how they cause digital images to blur. Having developed a “tool” based on a spreadsheet, students can manipulate the convolution kernel directly resulting in a statistically significant improvement in their performance on convolution questions.
Proposal Track
Non-research Project
Start Date
3-7-2014 2:45 PM
End Date
3-7-2014 3:30 PM
Recommended Citation
Rowan, Jim, "A Spreadsheet-Based Manipulative to Support Teaching Convolution" (2014). Interdisciplinary STEM Teaching & Learning Conference (2012-2019). 40.
https://digitalcommons.georgiasouthern.edu/stem/2014/2014/40
A Spreadsheet-Based Manipulative to Support Teaching Convolution
Room 2904 A
Blurring an image is fairly common but the question is, how does one blur a digital picture? The simple answer is to apply a blur filter…but how does that blur filter blur a picture that is stored as numbers? If it were a physical charcoal drawing the artist would smear the charcoal out with his finger to transform a sharp, dark line into a wider, lighter gray line but your camera produces images that are collections of numbers. How does that work?
Convolution calculations underlie many of the filters found in bitmapped image processing programs like GIMP and Photoshop. In Georgia Gwinnett College’s Digital Media class we “peek under the covers” to see how things work because we believe that knowing a little about what is going on helps dispel the idea that it is all magic performed by wizards. In the process of doing that the hope is to empower the student by demonstrating that he can understand what is going on and is not simply a slave to the gods of computing. One of the “under the cover peeks” we take is to look at these convolution calculations and how they cause digital images to blur. Having developed a “tool” based on a spreadsheet, students can manipulate the convolution kernel directly resulting in a statistically significant improvement in their performance on convolution questions.
