2009 Vol. 2, No. 5
2009, 2(5): 364-376.
Abstract
PDF 2037KB
Abstract:
In order to meet on going requirement of information for exploration to the earth and planet, space remote sensor must be manufactured with high spatial resolution, high spectrum resolution, high radiation resolution and super imagine quality while keeping light weight, besides that, the adaptability of space remote sensor to the variety rigor launching and space vacuum thermal environment must be taken into consideration. It is depicted summarily that the present state of environment adaptability design and the influence of mechanics and vacuum thermal environment on the performance of space remote sensor, and probe into the mechanism of environment effect, and the key points of environment adaptability design. The environment adaptability factors must be taken into account as a whole in the earlier design phase. It is pointed out that mass character; dynamical behavior and thermal optics character of space optical remote sensor are the most important factors to be considered in the course of iteration of design and analysis. During the production engineering stage, environment adaptability test should be carried out for assessing practice dynamical behavior and thermal optics character of product, and verify the validity of performance of space optical remote sensor for the sake of insuring sensor’s adaptability against variety rigor environment. The appraise method of environment adaptability design and the related test content are also described in this paper.
In order to meet on going requirement of information for exploration to the earth and planet, space remote sensor must be manufactured with high spatial resolution, high spectrum resolution, high radiation resolution and super imagine quality while keeping light weight, besides that, the adaptability of space remote sensor to the variety rigor launching and space vacuum thermal environment must be taken into consideration. It is depicted summarily that the present state of environment adaptability design and the influence of mechanics and vacuum thermal environment on the performance of space remote sensor, and probe into the mechanism of environment effect, and the key points of environment adaptability design. The environment adaptability factors must be taken into account as a whole in the earlier design phase. It is pointed out that mass character; dynamical behavior and thermal optics character of space optical remote sensor are the most important factors to be considered in the course of iteration of design and analysis. During the production engineering stage, environment adaptability test should be carried out for assessing practice dynamical behavior and thermal optics character of product, and verify the validity of performance of space optical remote sensor for the sake of insuring sensor’s adaptability against variety rigor environment. The appraise method of environment adaptability design and the related test content are also described in this paper.
2009, 2(5): 377-387.
Abstract:
Face recognition is an important problem of pattern recognition and machine learning. Among many approaches to the problem of face recognition, subspace analysis gives the most promising results, and becomes one of the most popular methods. This paper researches subspace analysis methods, introduces the basic theory of linear subspace such as PCA、LDA、ICA 、FastICA etc. and non-linear subspace such as KPCA etc. and their application in face recognition ,including some new research fruits concretely . In addition ,ORL database and YALE B database are used to experiment basic subspace methods. The experiment results indicate that FastICA method is more powerful than other subspace methods for face recognition. Finally, the advantage and disadvantage of these methods are discussed by the experiment results.
Face recognition is an important problem of pattern recognition and machine learning. Among many approaches to the problem of face recognition, subspace analysis gives the most promising results, and becomes one of the most popular methods. This paper researches subspace analysis methods, introduces the basic theory of linear subspace such as PCA、LDA、ICA 、FastICA etc. and non-linear subspace such as KPCA etc. and their application in face recognition ,including some new research fruits concretely . In addition ,ORL database and YALE B database are used to experiment basic subspace methods. The experiment results indicate that FastICA method is more powerful than other subspace methods for face recognition. Finally, the advantage and disadvantage of these methods are discussed by the experiment results.
2009, 2(5): 388-394.
Abstract:
Object tracking is a crucial subject in computer vision. Due to the variety of the target feature and the complexity of the background, target tracking has been a challenging subject all the time. Especially the tracking under occlusions, is one of the difficulties. How to deal with tracking under occlusion effectively, especially tracking under severely occlusion and totally occlusion, becomes the crucial restrictive factor to the robustness of tracking algorithm. This paper summarizes and classifies the main algorithms of tracking under occlusions home and abroad, expounds the basic ideas of these algorithms, introduces briefly the representative algorithms, analyses the advantages and disadvantages of these algorithms, and the situation that the algorithms fit. At last, necessary discussion is given about tracking under occlusions.
Object tracking is a crucial subject in computer vision. Due to the variety of the target feature and the complexity of the background, target tracking has been a challenging subject all the time. Especially the tracking under occlusions, is one of the difficulties. How to deal with tracking under occlusion effectively, especially tracking under severely occlusion and totally occlusion, becomes the crucial restrictive factor to the robustness of tracking algorithm. This paper summarizes and classifies the main algorithms of tracking under occlusions home and abroad, expounds the basic ideas of these algorithms, introduces briefly the representative algorithms, analyses the advantages and disadvantages of these algorithms, and the situation that the algorithms fit. At last, necessary discussion is given about tracking under occlusions.
2009, 2(5): 395-401.
Abstract:
Image enhancement is the most attractive areas in image-processing. In the embedded video image processing system, there are two key issues, quality of the enhanced image and speed of the image-processing, which decide the efficiency of the algorithm. And, the solution to the issues is the reasonable connection of appropriate image enhancement algorithm and efficient hardware platform. Depending on studying many typical enhancement algorithms，simulate three typical enhancement algorithm on matlab platform. The algorithms are histogram equalization, Homomorphic filtering, wavelet. Finally, analyze the simulation results and compare the advantages and disadvantages of each algorithm, point out their best scenes for application and the most efficient implementation hardware platform. In order to conclude a set of effective application instructing rules.
Image enhancement is the most attractive areas in image-processing. In the embedded video image processing system, there are two key issues, quality of the enhanced image and speed of the image-processing, which decide the efficiency of the algorithm. And, the solution to the issues is the reasonable connection of appropriate image enhancement algorithm and efficient hardware platform. Depending on studying many typical enhancement algorithms，simulate three typical enhancement algorithm on matlab platform. The algorithms are histogram equalization, Homomorphic filtering, wavelet. Finally, analyze the simulation results and compare the advantages and disadvantages of each algorithm, point out their best scenes for application and the most efficient implementation hardware platform. In order to conclude a set of effective application instructing rules.
2009, 2(5): 402-407.
Abstract:
The stray light analysis of a diffractive/refractive long wavelength infrared (LWIR) staring imaging system is shown. In this LWIR system, there is one diffractive optical element(DOE) fabricated by the diamond turning technology. In this paper, three sources of the stray light- unwanted diffractive order light of the DOE, the multiple reflected light of the optical surface, and the reflected light from the inner side of the barrel- are analyzed with the help of LIGHTTOOLS software. From the simulation result of 6 twice-reflections, it is shown that when the source is moralized, which results in the irradiance of 100 W/mm2 at the image plane, the twice-reflection result in 0.01 W/mm2 at the image plane; the inner side of the barrel, which has 10％ reflection, result in 0.01 W/mm2 at the image plane. The analyzed result is approved by the image get from the LWIR system.
The stray light analysis of a diffractive/refractive long wavelength infrared (LWIR) staring imaging system is shown. In this LWIR system, there is one diffractive optical element(DOE) fabricated by the diamond turning technology. In this paper, three sources of the stray light- unwanted diffractive order light of the DOE, the multiple reflected light of the optical surface, and the reflected light from the inner side of the barrel- are analyzed with the help of LIGHTTOOLS software. From the simulation result of 6 twice-reflections, it is shown that when the source is moralized, which results in the irradiance of 100 W/mm2 at the image plane, the twice-reflection result in 0.01 W/mm2 at the image plane; the inner side of the barrel, which has 10％ reflection, result in 0.01 W/mm2 at the image plane. The analyzed result is approved by the image get from the LWIR system.
2009, 2(5): 408-413.
Abstract:
The principle of coronagraph requires that stray light in coronagraph system must be kept at an extremely low level. Based on the principle of mirror coronagraph, the internally occulted mirror coronagraph is designed by analyzing its optical character and the basic rules of suppressing the stay light. In this system, the FOV is 0.67degree, the aperture is 47mm, the focal length is 768mm, the total length is 1200mm, the MTF value at 30lp/mm is above 0.6, the radius of spot is less than 2.5 um,and the image quality of the system approaches the diffraction limit.Considering the character of stray light in system,the structure of suppressing stray light is built.In the system,the main sources of stray light can be suppressed entirely .The system can achieve a level of 10-6-10-8B⊙ approximately in suppressing stray light,and so it can get high quality coronal images.
The principle of coronagraph requires that stray light in coronagraph system must be kept at an extremely low level. Based on the principle of mirror coronagraph, the internally occulted mirror coronagraph is designed by analyzing its optical character and the basic rules of suppressing the stay light. In this system, the FOV is 0.67degree, the aperture is 47mm, the focal length is 768mm, the total length is 1200mm, the MTF value at 30lp/mm is above 0.6, the radius of spot is less than 2.5 um,and the image quality of the system approaches the diffraction limit.Considering the character of stray light in system,the structure of suppressing stray light is built.In the system,the main sources of stray light can be suppressed entirely .The system can achieve a level of 10-6-10-8B⊙ approximately in suppressing stray light,and so it can get high quality coronal images.
2009, 2(5): 414-420.
Abstract:
In order to improve the zone polishing technology, material removal theoretical model of zone polishing technology is founded based on Preston equation. Considering all influencing factors, such as rotating speed ratio, eccentric distance, pressure distribution between polishing lap and workpiece, the mathematical model which is about the relationship between the material removal and the influencing factors is established. The theoretical analysis and experiment results show that: material removal efficiency is increased as the rotating speed ratio and eccentric distance increase. When the rotating speed ratio is near to one, the removal distribution is more uniform. The material removal efficiency of edge is quickly reduced as the edge of workpiece holds out of polishing lap. The zone polishing technology is improved through researching the technological parameters of the theoretical model, and the polishing efficiency and stability are increased.
In order to improve the zone polishing technology, material removal theoretical model of zone polishing technology is founded based on Preston equation. Considering all influencing factors, such as rotating speed ratio, eccentric distance, pressure distribution between polishing lap and workpiece, the mathematical model which is about the relationship between the material removal and the influencing factors is established. The theoretical analysis and experiment results show that: material removal efficiency is increased as the rotating speed ratio and eccentric distance increase. When the rotating speed ratio is near to one, the removal distribution is more uniform. The material removal efficiency of edge is quickly reduced as the edge of workpiece holds out of polishing lap. The zone polishing technology is improved through researching the technological parameters of the theoretical model, and the polishing efficiency and stability are increased.
2009, 2(5): 421-425.
Abstract:
A phase mask interferometer is developed to photo-write gratings with arbitrary Bragg wavelength. In this system, the gratings are inscribed by the UV interference fringes derived from two rotatable prisms, where the phase mask is used as a beam splitter. Furthermore, when the undersides of two prisms with the vertex angle defined by the #61617;1 diffraction angle of phase mask and the refract index of the prism are parapllel each other, the phase mask initializes the reference quantity of Bragg wavelength. As the shift of Bragg wavelength is 1 nm, the maximum rotation angle of the prism is ~1 degree, and the minimum rotation angle is ~2.4 minute. By contrasting with the rotation angle ~23 second/nm of the mirror in Talbot interferometer, the rotation precision of the prisms is decreased by two or three orders of magnitude in this phase mask interferometer. Instead of many phase masks with different grating periods, the phase mask interferometer writes grating with the inscribed wavelength of 1450~1600 nm via a phase mask and two rotatable prisms.
A phase mask interferometer is developed to photo-write gratings with arbitrary Bragg wavelength. In this system, the gratings are inscribed by the UV interference fringes derived from two rotatable prisms, where the phase mask is used as a beam splitter. Furthermore, when the undersides of two prisms with the vertex angle defined by the #61617;1 diffraction angle of phase mask and the refract index of the prism are parapllel each other, the phase mask initializes the reference quantity of Bragg wavelength. As the shift of Bragg wavelength is 1 nm, the maximum rotation angle of the prism is ~1 degree, and the minimum rotation angle is ~2.4 minute. By contrasting with the rotation angle ~23 second/nm of the mirror in Talbot interferometer, the rotation precision of the prisms is decreased by two or three orders of magnitude in this phase mask interferometer. Instead of many phase masks with different grating periods, the phase mask interferometer writes grating with the inscribed wavelength of 1450~1600 nm via a phase mask and two rotatable prisms.
2009, 2(5): 426-431.
Abstract:
The process of the acid denaturation of the albumin was studied using dynamic laser speckle methods in this work. The dynamic speckle pattern sequences were generated using the dynamic speckle theory firstly. And then, the acid denaturation process of the albumin was analyzed using the time history speckle patterns (THSP) and its gray level co-occurrence matrix (GLCM) obtained from simulation and experiments. Furthermore, this denaturation process was investigated throughly using the inertia moment of the GLCM and the mean contrast methods. The theoretical simulation and the experiments results show that the protein particles movement changed from high activity to low activity gradually in the colloid solution during the denaturation process, meanwhile, the quantity of the protein particles decreased and its size increased. The dynamic laser speckle method is a novel real-time, quick and effective method to study the albumin denaturation.
The process of the acid denaturation of the albumin was studied using dynamic laser speckle methods in this work. The dynamic speckle pattern sequences were generated using the dynamic speckle theory firstly. And then, the acid denaturation process of the albumin was analyzed using the time history speckle patterns (THSP) and its gray level co-occurrence matrix (GLCM) obtained from simulation and experiments. Furthermore, this denaturation process was investigated throughly using the inertia moment of the GLCM and the mean contrast methods. The theoretical simulation and the experiments results show that the protein particles movement changed from high activity to low activity gradually in the colloid solution during the denaturation process, meanwhile, the quantity of the protein particles decreased and its size increased. The dynamic laser speckle method is a novel real-time, quick and effective method to study the albumin denaturation.
2009, 2(5): 432-437.
Abstract:
Aiming at implementing loading custom code automatically when the TMS320C6416 DSP runs offline in the Photoelectrical and real-time Tracking system，this paper adopts the way of ROM Bootloader. But in default mode，it can only copy 1KB data. It is necessary to write the secondary bootloader programmes so that the rest custom code in external ROM/FLASH could be copied into the high speed of DSP. In this paper the detailed secondary bootloader programmes are written detailedly in the Non-BIOS and the BIOS system. The experiments show the method described in this paper achieved complete success. The Bootloader based on TMS320C6416 is well realized and thus the design requirement on the real-time and reliable ability of the Photoelectrical and real-time Tracking System can be satisfied.
Aiming at implementing loading custom code automatically when the TMS320C6416 DSP runs offline in the Photoelectrical and real-time Tracking system，this paper adopts the way of ROM Bootloader. But in default mode，it can only copy 1KB data. It is necessary to write the secondary bootloader programmes so that the rest custom code in external ROM/FLASH could be copied into the high speed of DSP. In this paper the detailed secondary bootloader programmes are written detailedly in the Non-BIOS and the BIOS system. The experiments show the method described in this paper achieved complete success. The Bootloader based on TMS320C6416 is well realized and thus the design requirement on the real-time and reliable ability of the Photoelectrical and real-time Tracking System can be satisfied.
2009, 2(5): 438-444.
Abstract:
According to request of data transfer in three-dimensional mapping camera, a new solution that will be more useful is investigated and developed. The image data in Multi-channe are combined with FPGA, the data is first cashed in FPGA in order to transfer the Asynchronous data into synchronous data. Then it will again be cached in FIFO. After these processes the data will be transferred to computer bus with PCI9054. With the design of driver and application, the image data can be showed and stored.
According to request of data transfer in three-dimensional mapping camera, a new solution that will be more useful is investigated and developed. The image data in Multi-channe are combined with FPGA, the data is first cashed in FPGA in order to transfer the Asynchronous data into synchronous data. Then it will again be cached in FIFO. After these processes the data will be transferred to computer bus with PCI9054. With the design of driver and application, the image data can be showed and stored.
2009, 2(5): 445-451.
Abstract:
According to image degeneration on foggy days, a penetrating fog imaging method using near infrared imaging and video image processing is proposed. According to the characteristics of low contrast image on foggy days and high comparability of histogram of contiguous field in video image, we have proposed an improved histogram equalization algorithm.This algorithm simplifies the system architecture as well as reduces the complexity of logical design. Moreover it also reduces the hardware resources of high speed memory. It has been simulated and verified on MATLAB and implemented on the video image processing platform based on FPGA. The experiment shows that the near infrared wave band has a better penetrating ability on fog. The improved algorithm enhances the image quality on foggy days while meet the real-time requirement for currently standard video image of 25 frame/s. With these two methods, we increase the visibility distance over 1.5 times on foggy days.
According to image degeneration on foggy days, a penetrating fog imaging method using near infrared imaging and video image processing is proposed. According to the characteristics of low contrast image on foggy days and high comparability of histogram of contiguous field in video image, we have proposed an improved histogram equalization algorithm.This algorithm simplifies the system architecture as well as reduces the complexity of logical design. Moreover it also reduces the hardware resources of high speed memory. It has been simulated and verified on MATLAB and implemented on the video image processing platform based on FPGA. The experiment shows that the near infrared wave band has a better penetrating ability on fog. The improved algorithm enhances the image quality on foggy days while meet the real-time requirement for currently standard video image of 25 frame/s. With these two methods, we increase the visibility distance over 1.5 times on foggy days.
2009, 2(5): 452-459.
Abstract:
Nonlinear filtering is an effective noise restraining technology and has been widely used. In order to solve the problem that the speed of classical median filter was slow because of a lot of sorting, in this paper, the method of the C code optimization based on the TMS320C6000 Series DSP is used to optimize the code of median filter. Compares with the original code, the optimized code reduces the running time, enhancing the performance of code execution. In addition, it will meet the requirement of real-time image processing.
Nonlinear filtering is an effective noise restraining technology and has been widely used. In order to solve the problem that the speed of classical median filter was slow because of a lot of sorting, in this paper, the method of the C code optimization based on the TMS320C6000 Series DSP is used to optimize the code of median filter. Compares with the original code, the optimized code reduces the running time, enhancing the performance of code execution. In addition, it will meet the requirement of real-time image processing.
2009, 2(5): 460-465.
Abstract:
A method based on the combination of input plane Mathematical Morphology processing and power plane FAF processing is proposed to improve the recognition ability of JTC. First, theory of classical JTC is introduced. Second, aiming at its disadvantage, the combination of input plane morphology processing and power spectrum plane FAF processing is used to improve the correlation output performance of JTC. Finally, The parameter of the proposed JTC are analyzed. Experimental results show that when the correlation peak is very sharpness. The peak value is , which is higher than classical JTC. Performance of the combination based method is better than that of any single method based JTC. Target detection ability is also improved by using the proposed JTC.
A method based on the combination of input plane Mathematical Morphology processing and power plane FAF processing is proposed to improve the recognition ability of JTC. First, theory of classical JTC is introduced. Second, aiming at its disadvantage, the combination of input plane morphology processing and power spectrum plane FAF processing is used to improve the correlation output performance of JTC. Finally, The parameter of the proposed JTC are analyzed. Experimental results show that when the correlation peak is very sharpness. The peak value is , which is higher than classical JTC. Performance of the combination based method is better than that of any single method based JTC. Target detection ability is also improved by using the proposed JTC.
