中文摘要：

为了实现对传统指针式化工仪表的数字化改造，提出了一种基于嵌入式机器视觉技术及物联网技术的自动读数方案。根据指针式仪表图像的特点，利用角度约束方法对霍夫变换进行了优化；并结合二值图像减影法，识别出指针偏转角度，从而将其转化成实际读数：最后利用无线传输实现远程自动抄表。实测结果表明，采用嵌入式机器视觉及物联网技术实现指针式化工仪表远程抄表具有较高的实用性、正确性及可行性。

英文摘要：

In order to realize the digital transformation for traditional chemical pointer type instruments, the automatic meter reading strategy based on the technologies of embedded machine vision and the Internet of things is proposed. In accordance with the features of pointer type instruments, by adopting angle constraint method, the Hough transformation is optimized, and combining with binary image subtraction algorithm, the deflection angle of pointer is identified and converted into actual reading, then remote automatic meter reading is implemented by using wireless transmission. The test results verify that using embedded machine vision and Interuet of things technologies to realize remote meter reading for the pointer type chemical instrumentation possesses higher practicability.