Against the backdrop of continuously accelerating software iteration speeds and increasing diversification of terminal devices, traditional automated testing tools that rely on control attributes struggle to meet requirements such as visual consistency and cross-device adaptation. Computer vision algorithms, with their capabilities in pixel-level analysis, accurate target localization, and dynamic scene recognition, have become a key technology to break through the "visual blind spots" in software testing. This paper sorts out the development trends of software test automation, analyzes the principles of core computer vision algorithms, expounds on application practices in scenarios such as UI (User Interface) verification and compatibility testing, examines current technical challenges, and proposes optimization paths. It aims to provide references for building a more comprehensive software test automation system.

Yuke Xing. Application of Computer Vision Algorithms in Software Test Automation. Academic Journal of Computing & Information Science (2025), Vol. 8, Issue 11: 116-121. https://doi.org/10.25236/AJCIS.2025.081113.

[1] Cui, Jing. (2025). Study on the Automated Testing for Computer Application Software based on Deep Learning. China-Arab States Science and Technology Forum (Chinese & English), (06), 97-100.

[2] Wang, Peng. (2025). Research on the Application of Artificial Intelligence in Computer Application Software Development. Mobile Information, 47(07), 335-337.

[3] Yin, Jie. (2024). Analysis of Automated Development Technology for Computer Application Software. Information Recording Materials, 25(03), 165-167.

[4] Li, Liang. (2025). An Automated Assembly Line Quality Inspection System Based on Computer Vision. Modern Manufacturing Technology and Equipment 61(06), 211-213, 218.

[5] Wang, Zheng. (2025). Research on Intelligent Testing and Optimization Technology of Artificial Intelligence in Computer Software Development. Information Recording Materials, 26(9), 30-32.