基于图像识别的航空滑油喷嘴流量试验台油位检测系统

考虑到使用静态容积法的航空滑油喷嘴流量试验台采用的磁致伸缩液位传感器在常温下具有较高的温度，但是温度高于常温时，会产生极大温度漂移无法保证试验系统的测量精度，本文将图像识别应用于航空滑油喷嘴流量测量系统中，对从喷嘴排出滑油体积进行自动识别，忽略掉温度变化、传感器精度等对测量精度的影响。通过图像滤波、分割、刻度数字识别、刻度线识别以及油位识别实现对油位的自动测量。通过实验方法与人工读数、磁致伸缩液位传感器读数以及温度补偿后的传感器读数进行对比。结果表明，使用图像识别技术的流量试验台的试验结果与真实值最为接近，能够有效提高检测的精度和效率，降低试验台成本和测试人员的工作量，其试验结果最为可靠。

Aviation lubricating oil nozzle flow test bench level detection system based on image recognition

Considering the slip oil nozzle flow test rig using the static volumetric method of aviation of magnetostrictive liquid level sensor at room temperature with high temperature, but the temperature is higher than room temperature, will produce great temperature drift can not guarantee the measurement precision of the test system. In this paper, the image recognition is applied in aviation lubricating oil nozzle flow measurement system, is discharged from the nozzle slide oil volume to carry on the automatic recognition of the neglected influence of temperature changes, and the sensor precision on the measurement accuracy. Through image filtering, segmentation, recognition, digital calibration scale line recognition and recognition level automatic measurement of oil level. By comparing the experimental method and the artificial reading, the magnetostrictive liquid level sensor and the temperature compensated sensor, the sensor readings were compared. The results show that the test results of the flow test bench with the image recognition technology are the most close to the real value, which can effectively improve the accuracy and efficiency of the test, reduce the cost of the test bed and the workload of the test staff, and the results are most reliable.