Research On Alignment Detection Method For Blackbody Cavity Sensor

Blackbody cavity temperature sensor has achieved high precision,low cost and the continuous temperature measurement.The response speed is the most important technical factor.The experiment proved that the alignment of the sensor and the nozzle has a significant impact on speed.At present,there are few studies on the alignment of long-axis small-aperture blind-hole structure.By use of computer vision detection technology,it can effectively improve the measurement efficiency and accuracy.The article analyzes the development of alignment problem and machine vision.It designs a method on alignment problem based on machine vision.According to the requirement of research,to design the neutral detection system in laboratory,which lay the foundation for the research and application of the method.To set up the neutral detection system in the laboratory,and to introduce the structure and experimental methods of the system,and to design different algorithms to detect different errors.Because of the position of plane is unchanged,the ellipse fitting method is designed to detect the translation error.To use Lambert's reflection theory and analyze the reflection model,the gray-level peak analysis is designed to detect the rotation error.To analyze image characteristics and to design the image processing method which is suitable for the object extraction feature.Through Gauss surface fitting,the edge coordinates are extracted in sub-pixels level.Calibration for the image size.To find the length of the corresponding relationship between physical and pixel.In this article,we obtained a satisfactory alignment detection algorithm by a large number of experimental algorithm studies.It achieves a 5%detection accuracy in position offset,and a 0.15° detection accuracy in angle offset.