Image Quality Detection Of Optical Fiber Image Transmission Bundle

Optical fiber image beam is a passive device that uses a certain number of regularly arranged optical fibers to realize image transmission.It plays an important role in the fields of medicine,military,aerospace,scientific research and so on.However,in the manufacturing process,due to the processing error or low equipment accuracy,the optical fiber image beam often has problems such as broken wire,dark wire or reduced image resolution.Therefore,it is necessary to detect the imaging quality of optical fiber image beam.Firstly,this paper studies the image detection technology in related fields and its development at this stage,and then designs and implements the image acquisition system to complete the independent acquisition of the end face image of the optical fiber image beam;According to the structure and characteristics of the end face image,the algorithms such as brightness equalization compensation,adaptive contrast enhancement,filter denoising,morphological operation and edge detection are used for image processing,and the unified computing equipment architecture is used to accelerate the optimization of the algorithm and improve the work efficiency;On this basis,the detection of the number of pixels and the bad point rate of the optical fiber image beam are studied respectively,and the edge optimization processing algorithm based on Euler number calculation and the tree marking method based on multi-layer area threshold are proposed to realize the number detection and defect classification;Finally,by adding lens and changing optical path,the comprehensive resolution detection experiment of optical fiber image beam imaging system is completed,and the scheme and design to improve the imaging quality of the system are studied.The experimental results show that the algorithm used in this paper has high applicability to all stages of image processing of the end face of the optical fiber image beam.On this basis,the detection error rate of the number of pixels is about 0.271%,which meets the design requirements;The defect detection is marked according to the "one-to-one" principle,and the bad point rate of the sample is calculated to be about 0.833%,which meets the qualified index;The comprehensive resolution of the system obtained by the experimental detection is basically consistent with the theoretical calculation value of the limit resolution of the optical fiber image beam,which proves the feasibility of the detection scheme.