Scintillation Crystal Growth Quality Evaluation Based On Machine Learning

The scintillator crystal is a crystalline material that can light under the influence of high-energy rays and be widely used for detecting particles. With development industrial CT other fields, demand inorganic crystals with high melting points performance increasing yearly, which results in abandonment traditional laboratory mode manufacturing crystals. For achieving needs high-standard crystals, market attaches great importance to intelligent One critical technical difficulties numerical assessment quality. research relies on manually measuring quality, has many defects, such as inability numerical, quality evaluation varies from person person. Therefore, this paper proposes methodology assessing based residual depth network. At same time, order make more refined, uses target detection method learning determine area. model article developed yolov5 deep framework, suitable detection, its accuracy 98%. network proposed by ResNet-18, grading, reaches 84.8%. scintillation solution problem representation opens up channel closed-loop control processes.