Research On Auto Focus Technology In Optical Imaging System

As one of the main methods of space observation and aerial detection,optical imaging system such as telescope and aerial camera is an important approach to space targets perception and aerial target detection.Due to its advantages of high-resolution and strong real-time,the telescope system has been widely applied in space information detection and air defense.However,the actual application environment of the optical imaging system is complex and changeable.Various factors such as temperature,pressure,impact,friction,etc.may cause defocusing in the imaging process of such a sophisticated and complex instrument,which will cause the image quality to deteriorate and affect subsequent work.Focusing technology is an important method to improve the imaging quality of optical system,and defocusing detection technology is the prerequisite and foundation for the accurate implementation of focusing.Therefore,fast and high-precision defocusing detection methods are important research contents of high-quality optical imaging systems.By studying various types of defocusing detection technologies,analyzing their development history,testing principles,and comparing their advantages and disadvantages,this thesis deeply discusses the sensor-based aberration detection technology and the defocusing detection method based on image processing.On this basis,a method for improving the real-time and accuracy of the extended target is proposed.First of all,a detailed summary and explanation of the research background,research significance,various technical methods and development,basic principles and models of defocusing are proposed in this thesis.In order to achieve the defocusing detection in the point target,this thesis has completed the detection of the defocusing amount based on the research of the ShackHartman sensor,with an average error of 0.03?,and analyzes various factors that affect its accuracy.In order to improve the speed of defocusing detection on the extended target,this thesis proposes a method of directly processing the spectrum of the defocused image on the basis of the focusing method based on image processing.First,the blur parameters are directly obtained according to the frequency domain characteristics of the defocused image,and then the coarse defocusing detection of the extended target can be completed.This method can save a lot of adjustment time in the image-based detection and focus system,so the time for adjustment,transmission,and detection of multiple images are shortened to about 1s.In order to improve the accuracy of defocusing detection on the extended target,an improved sharpness evaluation function based on image gray gradient operator is proposed.By enriching the gray gradient direction of the operator,optimizing the threshold and weight,this function improves an accuracy about 0.0855 to 0.2145 compared with the traditional evaluation function,and greatly improves the anti-noise performance.In the verification of the real defocused extended target,the sensitivity factor of this method showed an increase of at least 13% and 17% respectively in a single scene and a complex scene.Finally,a fast and high-precision defocusing detection method combining coarse focusing and fine focusing is proposed for the aerial extended target,which provides a guarantee for high-quality images of aerial target detection.