Research On Image Background Noise Elimination Method Based On Feature Spectrum Imaging Technology

The recognition of small and dim targets in the sky background is all that matters to the flight safety of spacecraft.At present,the methods of target recognition mainly include radar detection and infrared recognition,etc.However,radar detection is vulnerable to electromagnetic interference,and the signal-to-noise ratio of infrared recognition is low.There are still shortcomings in areas that require high accuracy,such as target recognition and aerospace security.The feature spectrum elimination of background noise technology is a new method of identification and positioning with strong antiinterference ability and real-time performance.It has become a research hotspot in spectral imaging and digital image processing.According to the application requirements of using characteristic spectra to eliminate or suppress background noise,and considering the differences in the spectra of different atoms during combustion,this thesis analyzes the flying object combustion composition and the flying object tail flame spectrum,and combines the application requirements to design and manufacture a dedicated micro/nano filter.Experiments have studied the methods to eliminate the characteristic spectral background noise in different environments,and achieved high-quality imaging and contrast enhancement.The main work of this paper includes:(1)A method is proposed that based on two different wavelength filters after image subtraction to eliminate background noise and its effectiveness is verified by experiments.According to the characteristics of the atomic spectrum in the tail flame of the flying object,The 589 nm of sodium atom radiation was selected as the characteristic wavelength of the characteristic spectral imaging experiment.In the experiment,the narrowband filter with a center wavelength of 589 nm and 579 nm was placed in front of the camera.The camera gets two pictures,one with the target and background,and the other with the background.After processing the image gray scale,feature matching,etc.,the two images are subtracted,and the background noise image can be eliminated.In infield experiments and field test respectively,the results show that the contrast between the target and the background can be increased by 19 times and 131 times by using this method.(2)A special portable four-channel micro/nano filter was designed and manufacture.The filter can obtain four-channel optical grayscale images in one exposure shooting,and can eliminate the interference caused by position shift jitter and exposure differences between images.We designed a film that meets the test requirements,tested the etching rate of the film under different conditions,and taped out a micro/nano filter with a specific geometric size.After the filter is connected with the sensor,it can be equipped in an industrial camera to realize snapshot spectral imaging.Compared with traditional spectroscopic imaging technology,this method has obvious advantages.