Research On DKDP Crystal Damage Detection Based On Image Processing

Potassium dihydrogen phosphate(KDP)and potassium deuterium phosphate(DKDP)crystals have the advantages of large nonlinear optical coefficient and can grow large-aperture single crystals,and are often used as electro-optical switching elements and frequency conversion devices in high-power large-aperture laser devices.Under continuous laser irradiation,the crystal will be damaged,which is called Laser-induced Damage,LID).The occurrence of damage will affect the operation of the whole optical system,so it is necessary to deal with and analyze the damage in the crystal in time,and crystal damage is one of the reasons that restrict the output capacity of high-power laser devices.In KDP/DKDP crystals,the density of body damage points is one of the criteria to evaluate the quality,damage degree,shading rate,and whether the crystal can continue to be used.Analyzing the density and growth of body damage points can help to understand the damage mechanism and characteristics of the crystal.In this paper,based on image processing technology,the research on DKDP crystal damage detection is carried out,and the automatic detection of DKDP crystal body damage is completed from two parts: pre-imaging optimization and image algorithm.The main research is as follows:1.Build an online detection platform for DKDP crystal body damage.Aiming at the influence of imaging system on the accuracy of damage detection,a volume damage detection method based on scattering imaging is studied.The effects of different sources and camera gain on the morphology of damage points are analyzed.The causes of the birefringence of DKDP crystal are analyzed.Aiming at the misjudgment of the damage point caused by the birefringence,a polarizer is used to eliminate the virtual image in the crystal damage image to avoid repeatedly extracting the damage information at the same position.Through the above process,the image imaging can be optimized in the early stage and the image contrast can be improved.2.the image processing algorithm of DKDP crystal body damage detection is studied.Based on generating a signal image based on the Local Area Signal Strength Ratio(LASSR),this paper studies the generation of a high-frequency background suppression image based on an image difference algorithm and proposes an adaptive gray enhancement algorithm based on the local histogram for the possible missed detection of weak damaged targets in the image.In order to effectively segment the target,common classical image segmentation algorithms were studied and applied to segment the damaged regions and extract target features.To address the issue of noise in the segmentation,an iterative median filtering algorithm was utilized for elimination.Algorithms from morphological processing were investigated to tackle challenges such as adhesion,spurious points,and fractures in the damaged regions.To accomplish the final extraction of the damaged target,edge detection and edge tracing were performed on the segmented target.Boundary fitting was applied to extract connected domains,enabling the labeling of the damaged regions and the statistical analysis of their areas.The results indicate that this algorithmic process effectively resolves the recognition issue of weak damage points in damaged images and improves detection accuracy.3.design and complete the testing software for DKDP crystal body damage detection.The image processing algorithm of DKDP crystal body damage detection is integrated into the test software,and the test methods under different laser damage threshold conditions are designed.Experiments show that the system can be used in the laser damage threshold test system,with a single frame image detection speed of 0.81 s and a maximum detection accuracy of 3 pixels,and the online detection of DKDP crystal body damage is completed.