Space-based Background Target Acquisition Algorithm In The Application Of Multispectral Spectrometer

Space mission is an important technical mean for space analysis and design. Withthe aerospace products playing an increasingly important role in the military andcivilian field, space missions increasingly diverse and complex, space missionsimulation is playing an increasingly important role. Multi-spectral imagingspectrometer is the recent developed new remote sensing instrument. Throughobtaining the target’s two-dimensional full color image information, the instrumentachieves the tracking of the man-made celestial objects and obtains much spectralinformation about the object to complete the target recognition and classification.Through the two-dimensional full color image obtaining by multi-spectral imagingspectrometer, achieving target recognition and tracking of multi-target acquisitionalgorithm is becoming the core and key.So the target acquisition algorithm and the design of the algorithm’s simulationplatform is more meaningful.Based on the comparison of the existing recognition methods, this paper startsfrom analyzing the full-color image obtaining by multi-spectral imaging spectrometerand proposed improved algorithms——difference algorithm based on Fourier-Mellinimage registration and edge extracting; double-difference algorithm based onFourier-Mellin image registration; difference algorithm based on Fourier-Mellinimage registration and centroid registration.First, we introduce the flow of the difference algorithm based on Fourier-Mellinimage registration and edge extracting. Firstly, it uses Fourier-Mellin algorithm andphase correlation algorithm to obtain the rotation angle, scaling and translationparameters. So we can get the registration point. Then, we register the collected twocontinuous images, count the difference of the two images and binarize the twoimages. Lastely, we multiply the difference image with the after edge extractingimgage to obtain the targets.Second, we introduce the flow of the double-difference algorithm based onFourier-Mellin image registration. Firstly, it uses Fourier-Mellin algorithm and phasecorrelation algorithm to obtain the rotation angle, scaling and translation parameters.So we can get the registration point. Then, we register the collected three continuousimages, count the difference of the three images and binarize the three images. Lastly,we multiply the two difference image with the after edge extracting image to obtainthe targets.Last, we introduce the flow of the difference algorithm based on Fourier-Mellinimage registration and centroid registration. Firstly, we use phase correlation to getthe rotation angle. Then, we recognize the target.In this paper we use spartan-3a development board to set up the verificationplatform. We firstly use STK to get the images. Then, we send the data to theSpartan-3A DSP1800a through RS232.The main work of this paper is that we complete the research of the acquisitionalgorithm under the space-based background and set up the verification platformabout this algorithm. Also, we propose a few useful suggestions about further work.The main innovation of this paper is: proposed the improved detectionmethod—difference algorithm based on Fourier-Mellin image registration and edgeextracting; double-difference algorithm based on Fourier-Mellin image registration;difference algorithm based on Fourier-Mellin image registration and centroid registration. Complete the software design and hardware design about this acquisitionmethod under space-based background.