Measurement Method of Interpupillary Distance and Pupil Height Based on Ensemble of Regression Trees and the BlendMask Algorithm

Measuring interpupilary distance and pupil height is a crucial step in the process of optometry. However, existing methods suffer from low accuracy, high cost, lack portability, limited research on studying both parameters simultaneously. To overcome these challenges, we propose method that combines ensemble regression trees (ERT) with BlendMask algorithm to accurately measure interpupillary height. First, train an ERT-based face keypoint model locate pupils calculate their center coordinates. Then, develop eyeglass dataset obtain coordinates lowest point lenses. Finally, numerical values based respective definitions. The experimental results demonstrate proposed can (IPD) height, calculated IPD are good agreement measurements obtained by auto-refractometer. By combining advantages two models, our overcomes limitations traditional measurement strong portability. Moreover, this enables fast automatic measurement, minimizing operation time, reducing human errors. Therefore, it possesses broad prospects for application, particularly fields customization vision inspection.