Spot the difference: automating visual regression testing
summary
Claims that pixel comparisons add practical value and ships a custom C# visual-test framework that piggybacks on Selenium to take stitched and element-level screenshots, compare against base images in ~4s for a Google snapshot, and produce diff images with highlighted boxes. It finds image and style regressions that DOM checks miss, reduced some manual checks but revealed maintenance costs: suites can still take ~4 hours, per-browser base images and anti-aliasing tolerances are required, and the repo plus NuGet prove the idea.
AI-written summary · about
chapters (20)
- 0:00Start
- 1:26COMMON UI AUTOMATION ISSUES
- 7:00TESTING PYRAMD
- 7:27WHY NOT JUST STICK WITH MANUAL TESTING THEN
- 7:43MANUAL TESTING
- 9:39CATCH 22
- 12:10SELENIUM
- 22:07VISUAL TESTING
- 24:50ROLLING OUR OWN
- 25:59EXECUTION SPEED
- 27:14HUMAN VS MACHINE
- 28:54MAINTAINABILITY
- 30:47IMAGE QUALITY DIFFERENCE
- 31:04MAINTAINING BASE
- 31:34SCREENSHOTS
- 32:44Responsive design
- 32:54HARDWARE ISSUES
- 33:46STORAGE
- 35:13MOVING ELEMENTS
- 36:25PDF DYNAMIC CONTENT
related talks
37mSpot the Difference; Automating Visual Regression Testing
37mThe Colorful World of Visual Test Automation
27mVisual UI Testing – Testing the Last Mile
40mVisual regression testing: your strategy's blind spot
42mHow Visual AI Makes Testing Easy
31mAutomating UX Compliance Testing
18mImplementing a Visual CSS Testing Framework
12mIntegrating visual testing into the build pipeline