基于图像识别的微型倾斜传感器

通过人眼观察进行识读的传统水准器式倾斜仪表，直观但精度不高、灵活性不强。通过引入CCD图像传感器采集水准器中气泡图像并结合图像识别算法分析计算气泡图形边缘，研究开发了具备倾斜角度数值输出能力的新型倾斜传感器。并进一步通过采用DSP处理器并应用嵌入式设计方法，设计了该类型倾斜传感器的小型化原型，从而显著地提高了测量精度和应用的灵活性．克服了传统仪表的不足。

A Micro-Tile Sensor Based on Image Recognition

With traditional leveling survey type tilt instruments, the measurement is obtained by observation of human eye, so the accuracy is not high enough and the flexibility is low. A new type of tilt sensor with digital tilt data output is developed through introducing CCD image transducer to collect bubble image in leveling survey meter, as well as to analyze and calculate bubble image edge by combining image recognition algorithm. Then, the slope sensor prototype of small size and intelligence is designed by utilizing embedded methodology and DSP processor, which highly improves the measuring accuracy and flexibility of application and overcome the insufficiency.