CAJ | 학술논문

为了提高材料试验机测量值的读取效率、降低人工操作造成的误差和减轻操作员的工作强度，提出一种具有较强鲁棒性的实时材料试验机测量值自动识别方法，利用一系列图像处理技术对材料试验机测量值图像进行识别。实验证明识别准确率和识别速度完全满足实际需求。这里创新地对数字切分中处理断裂数字的合并和粘连数字的分割采用二次阈值化分割法。
위료제고재료시험궤측량치적독취효솔、강저인공조작조성적오차화감경조작원적공작강도，제출일충구유교강로봉성적실시재료시험궤측량치자동식별방법，이용일계렬도상처리기술대재료시험궤측량치도상진행식별。실험증명식별준학솔화식별속도완전만족실제수구。저리창신지대수자절분중처리단렬수자적합병화점련수자적분할채용이차역치화분할법。
In order to improve the reading efficiency of values measured by material testing machine,reduce human error caused by the operators and lighten the working intensity of the operators,a robust real-time automatic identification method of values measured by material testing machine is presented in this paper. A series of image processing technologies were adopted to identify the digital images measured by material testing machine. The experimental results show that the recognition accuracy and speed can fully meet the actual needs of the system. The innovation is to use the secondary threshold method to deal with the broken digit merger and adhent digit division.