Hyperspectral anomaly detection (HAD) is a technique to find observations without prior knowledge, which is of particular interest as a branch of remote sensing object detection. However, the application of HAD is limited by various challenges, such as high-dimensional data, high intraclass variability, redundant information, and limited samples. To overcome these restrictions, we report an unsupervised strategy to implement HAD by dimensionality reduction (DR) and prior-based collaborative representation with adaptive global salient weight. The proposed framework includes three main steps. First, we select the most discriminating bands as the input hyperspectral images for subsequent processing in a DR manner. Then, we apply piecewise-smooth prior and local salient prior to collaborative representation to produce the initial detection map. Finally, to generate the final detection map, a global adaptive salient map is applied to the initial anomaly map to further highlight anomalies. Most importantly, the experimental results show that the proposed method outperforms alternative detectors on several datasets over different scenes. In particular, on the Gulfport dataset, the area under the curve value obtained by the proposed method is 0.9932, which is higher than the second-best method, convolutional neural network detector, by 0.0071.
The existence of forgeries has seriously affected the fair trading, protection and inheritance of calligraphy and painting, while it has been unable to identify high-level counterfeiting means by traditional expert eye identification method. Combining the advantages of material attribute recognition and imaging analysis of hyperspectral imaging technology with the powerful feature expression and classification ability of convolutional neural network, the identification level of calligraphy and painting could be improved. However, there are still some practical problems in the application, like the small sample learning problem caused by the difficulty in obtaining the real hyperspectral sample data of calligraphy and painting. In this paper, a 10-hidden layers 2D-CNN convolutional neural network transfer learning method for calligraphy and painting identification with data enhancement is proposed by using a large number of relevant picture data and a small amount of MNF dimensionality reduced hyperspectral data. The experimental test shows that on the test set of this paper, for the identification of calligraphy and painting authors and authenticity, the accuracy of migration learning with data enhancement under the original sample are separately 97.5% and 94.8%, the accuracy of migration learning with data enhancement under half of the original sample are separately 94.3% and 92.8%, which shows the migration learning and data enhancement is helpful, and the identification accuracy of half of the original sample basically reaches the identification accuracy of the original sample without data enhancement and transfer learning, whose accuracy are 92.1% and 92.5%.
With the improvement of spatial resolution, the focal length of space cameras and spectral imagers become longer. The thermal stability of image stability is more sensitive, with the temperature, especially in VNIR (visible and near-infrared). To solve the thermal stability of R-C(Ritchey-Chrétien) long focal length fore-telescope system, the relevant factors are discussed, on the basis of the LASIS(Large Aperture Static Imaging Spectrometer), and the change in the spacing between primary mirror and secondary mirror with temperature is proposed. Base on the calculation in theory, the method of mechanical passive athermalization design is developed. Mechanical test results indicate that the first natural frequency is 195Hz, above the 100 Hz. The thermal experiments show that the stability of primary mirror and secondary mirror spacing is 0.5μm.℃-1 , consisting with the FEA(Finite Element Analysis) value
To address the problem that bonding can lead to a reduction in the surface shape precision of a space-bound mirror, relationships between mirror deformation, thermal stress, and curing shrinkage stress were studied, and a bonding microstress design route was proposed. The thermal stress and thermal deformation introduced by thermal expansion mismatch were eliminated through an athermal adhesive layer thickness design. The relationship between mirror deformation and the curing shrinkage of the adhesive layer was derived completely, and structural optimization measures for releasing the curing stress of the adhesive layer are given. Bonding stress analysis was conducted based on the equivalent thermal deformation method, and an optimal structure meeting the design requirements was obtained. Finally, bonding of the mirror assembly was completed via this route, and the measured surface shape precision was stable at 0.0225λ. The theoretical analysis and experimental study demonstrate that this bonding design method can predict the bonding stress in the assembly process, making the follow-up bonding result controllable. These results should provide an excellent reference for the design and high-precision integration of large-aperture mirrors.
Anomalous objects detection for hyperspectral imagery is a significant branch in the area of remote sensing. Although enormous advancements have been developed, issues of redundancy of spectral information and correlation between pixels should be further explored and improved. To address these problems, we proposed a method that is on the basis of integrating collaborative representation with multipurification processing and local salient weight. Multipurification processing consists of spectral bands purification (SBP) and background purification (BGP). First, to alleviate the interference of redundant spectral information, we remove unnecessary spectral bands by adopting SBP based on considering the global spectral intensity of each band. Then, we remove the outliers in the local dual window by BGP to avoid the effect of heterogeneous pixels. Simultaneously, we obtain the local salient weight by calculating the similarity and difference of pixels in the dual window. Next, we obtain the initial detection result by a collaborative representation, which has been testified to be very effective. Finally, combined with the local salient weight map, the initial detection map is improved to the final detection map. To demonstrate the superiority of the proposed method, we conducted the comprehensive experiment on three public benchmark datasets that contain 15 hyperspectral images.
In order to solve the problems of long manufacturing cycle and high processing cost of mirrors in reflective optical systems of space cameras, a method for manufacturing metal mirrors based on additive manufacturing process is proposed. This paper designs an open honeycomb structure on the mirror backplane, mirror blank is prepared by additive manufacturing technology. Preliminary improvement of surface quality with rough machining and diamond turning. Then, a high-precision mirror surface is obtained through surface modification and secondary diamond turning. The diameter of the prepared mirror is 110mm, mass reduction rate of 70% and surface shape accuracy is better than λ /15 RMS( λ =632.8nm).The results show that the metal mirror prepared by this process can meet the requirements of high-precision reflective optical systems. The research work in this article can provide technical reference for the application of additive manufacturing technology in the field of optics. It has important reference and guiding significance for the research and application of related fields.
The curved prisms have been widely used as a light splitting element because of its own focal power and can be set in nonparallel light path. When the effective light transmission area of large-size curved prism is rectangular or elliptical, it can be processed into rectangular optical elements during processing, which can reduce the volume and mass of the instrument on the one hand, and effectively block the stray light outside the field of view on the other hand. Adhesive fixation is one of the common fixation methods of optical elements, which is applied in many launched spaceborne remote sensors. However, the adhesive bonding process and adhesive strength are unstable and the bonding reliability is poor, resulting in the calculated theoretical bonding area and adhesive layer thickness are often difficult to meet the requirements of complex mechanical and thermal environment. The rigid-flexible dual mode coupling support structure for space-based rectangular curved prism was firstly introduced. And then the tensile and shear tests on the epoxy adhesive used in this project was carried out and the bonding area based on the strength test was designed. On this basis, the mechanical test of the simulator mirror group was carried out to verify the reliability of the bonding area and the design of the support structure. Finally, three bonding postures were simulated analysis and tests including prone, lateral and vertical bonding. The results showed that the vertical bonding was the smallest surface shape errors of the curved prism. Based on this bonding attitude, the bonding and mechanical tests of the curved prism were completed to verify the reliability and rationality of the bonding process.
Change detection is an important research direction in the field of remote sensing technology. However, for hyperspectral images, the nonlinear relationship between the two temporal images will increase the difficulty of judging whether the pixel is changed or not. To solve this problem, a hyperspectral change detection method is proposed in which the transformation matrices are obtained by using the constraint formula based on the minimum spectral angle, which uses both spectral and spatial information. Further, a kernel function is used to handle the nonlinear points. There are three main steps in the proposed method: first, the two temporal hyperspectral images are transformed into new dimensional space by a nonlinear function; second, in the dimension of observation, all the observations are combined into a vector, and then the two transformation matrices are obtained by using the formula of spectral angle constraint; and third, each pixel is given weight with a spatial weight map, which combined the spectral information and spatial information. Study results on three data sets indicate that the proposed method performs better than most unsupervised methods.
The research on optical-mechanical system based on additive manufacturing is based on additive manufacturing, diamond turning, high-precision magnetorheological polishing, surface modification technology, internal lattice structure topology optimization and so on. It can overcome the problems of adhesive use and material matching in the traditional structure design, greatly reduce the difficulty of assembly and thermal control, and realize the lightweight design of internal structure that can not be realized by traditional processing methods. In addition, because the mirror body and its supporting structure are made of same metal materials and integrated, the strength and stiffness are greatly improved compared with the traditional design method. This paper summarizes the development status and technical parameters of additive manufacturing opto-mechanical system at home and abroad. The research progress of surface modification technology by domestic and foreign scholars was focused and the post-processing process and core technologies of the optical-mechanical system were described.
In recent years, the commercialized low-cost rotor UAV equipped with small hyperspectral camera has become an emerging way to acquire hyperspectral remote sensing data due to its advantages of convenient data acquisition and low cost, and has been widely used in environmental monitoring, precision agriculture, ocean development and other fields. Due to its light weight, the flight process of the rotor UAV is vulnerable to the influence of air flow, which leads to the poor stability of the platform and resulting in the geometric distortion of the imaging. This situation is particularly prominent in the obvious air flow areas such as the lake and the sea, which restricts the application in related fields. In this paper, the geometric processing of UAV-borne hyperspectral measured data collected at a wharf in Qiandao Lake is studied. There are two major factors lead to geometric error. One is the longitude and latitude step error, the other is the frame frequency of the navigation data is lower than the image exposure frame frequency. In this situation, this paper proposes a step fitting method to perform geometric correction and error correction for the hyperspectral airstrip data. Compared with the traditional Kalman filtering method and the global linear fitting method, the proposed method can better correct the geometric distortion of hyperspectral image caused by the low-precision GPS/INS system, which provides a foundation for the subsequent quantitative application of lightweight UAV hyperspectral camera in various fields.
The data acquired by the space-borne interference hyperspectral imager is a hybrid interferogram cube. For this kind of complicated data, there are many compression schemes for the on-board compression encoder. How to determine the optimal compression scheme and the optimal compression parameters under this compression scheme is crucial to recovering data. This paper proposes three compression schemes for space-borne interference hyperspectral images, which are mixed interferograms, pure interferograms, and fast views, respectively, and performs a compression evaluation. The peak signal-to-noise ratio of the recovered image, minimum quadratic error, and the spectral angle corresponding to the restored spectral curve are used as the measurement indicators to determine the optimal scheme and the optimal compression parameter configuration which are successfully applied to the development of the spectrometer. This paper establishes a scene-rich remote sensing interference hyperspectral image data source, quantitatively evaluates the impact of different compression ratios under different compression schemes, and changes in remote sensing image displacement, guides instrument design and parameter configuration, and lays a good foundation for the data application of space-borne interference hyperspectral images.
In the process of wide-band spectrum detection, interferogram acquisition of the traditional Michelson interferometer needs to follow Nyquist sampling theorem, the static performance such as high resolution of moving mirror scanning and the dynamic performance such as transient response need to meet strict requirements, which usually make the spectrometer system structure complex. Meanwhile, the interference modulation efficiency of traditional Michelson interferometer will drop sharply with the increase of optical path difference(OPD). In this way, the interference data value at the long optical path difference will be submerged by noise, which will reduce the signal-to-noise ratio of reconstructed spectrum. In order to simultaneously achieve spectrum detection with wide-band spectrum, high resolution and high signal-to-noise ratio, this paper introduces a configuration of wide-band interference spectrometer based on band-pass sampling technology. The wide-band interference spectrometer includes dispersion unit and interference modulation unit. Firstly, the dispersion unit pre-disperses the wide spectrum into continuous spectrum distributed along wavelength and divides the interference modulation signal of continuous spectrum into several interference signals of narrow-band spectrum. Secondly, the interference modulation unit carries out interference modulation on the dispersed continuous spectrum and the interferograms of every narrow-band spectrum are sampled and obtain the interferogram sequence of every narrow-band spectrum according to the band-pass sampling theorem. Finally, the spectral distribution of the detection target can be obtained by data processing and spectral superposition. The interference spectrometer provides a new idea for the development of spectral detection with wide spectral range, high resolution and high signalto- noise ratio.
To ensure the high surface accuracy and high thermal stability of space mirror, a lightweight design for the Φ514mm ULE primary mirror of a space remote sensor and flexible support structure with three-point was carried out. By further optimizing the parameters of the flexible supporting structure, the requirements of the optical index were met. The finite element model of the mirror assembly was established, and the static and dynamic characteristics of the assembly were analyzed. The results showed that the surface shape accuracy (RMS) of the mirror assembly is better than 8 nm under a load case of 1g gravity when the optical axis is level, and the first-order natural frequency of the component is 254 Hz. Finally, a mechanical test was carried out on the mirror assembly. The test results showed that the first-order frequencies of the three directions of the mirror assembly are all greater than 100 Hz , the error between the test data and the finite element analysis results does not exceed 10%. Analysis and test results showed that, the reasonable support structure design can effectively lower the change of the mirror surface shape caused by assembly stress and thermal stress, and has good dynamic performance. It is verified that the mirror and its supporting structure designed in this paper are reasonable, which provides reference and ideas for the design of flexible supporting structure of similar space mirror.
Coastal environmental elements such as bathymetry maps are of great significance to the economic and military development of each country. Spaceborne hyperspectral imager is one of the important instruments for coastal zone monitoring. Firstly, this paper systematically reviews the index system of spaceborne hyperspectral imagers, and then introduces the applications of hyperspectral remote sensing images in the retrieval of nearshore bathymetry. In order to improve the inversion accuracy, the current research status and shortcomings of fusion technology of laser active remote sensing and hyperspectral passive remote sensing are discussed. Furthermore, the index system of hyperspectral imagers is prospected based on the requirement of applications in coastal zone monitoring, which provides reference and support for the further development of hyperspectral remote sensing in coastal applications.
To minimize the assembly stress and thermal stress introduced by the support structure, and ensure the high surface accuracy and high thermal stability of space mirror. An ultra-lightweight design of secondary mirror was carried out for some space remote sensor, and three tangential bipods were used for quasi-kinematic support. Firstly, the design principle of quasi-kinematic support structure was investigated, and advantages of bipod kinematic support were analyzed from the angle of degree of freedom decoupling. Based on structure designed above, the finite element model was established. Taking surface accuracy of the mirror as optimization objectives, the integrated optimization method was adopted to extract the structural parameters with high sensitivity on the surface accuracy in the flexible support structure, and parameters optimization design was carried out. Finally, the static and dynamic characteristics of the optimized mirror assembly were analyzed. The analysis results showed that the surface shape accuracy (RMS) of the mirror assembly is better than 1 nm under a load case of 1g gravity when the optical axis is level. Surface accuracy (RMS) is better than 2 nm under the load case of 4℃ uniform temperature rise. The first-order natural frequency of the secondary mirror assembly is 587 Hz. The optimized mirror support structure can well unload the additional deformation caused by the support structure, and has good dynamic stiffness, which verifies that the designed mirror and its support structure are reasonable, and the optimization design method is reliable. This paper provides a reference and idea for the design of flexible support structure of space mirror.
KEYWORDS: Image segmentation, Image processing algorithms and systems, Signal to noise ratio, Binary data, Cameras, Feature extraction, Lithium, Sensors, 3D vision, Stereo vision systems
Low-light stereo vision is a challenging problem because images captured in dark environment usually suffer from strong random noises. Some widely adopted algorithms, such as semiglobal matching, mainly depend on pixel-level information. The accuracy of local feature matching and disparity propagation decreases when pixels become noisy. Focusing on this problem, we proposed a matching algorithm that utilizes regional information to enhance the robustness to local noisy pixels. This algorithm is based on the framework of ADCensus feature and semiglobal matching. It extends the original algorithm in two ways. First, image segmentation information is added to solve the problem of incomplete path and improve the accuracy of cost calculation. Second, the matching cost volume is calculated with AD-SoftCensus measure that minimizes the impact of noise by changing the pattern of the census descriptor from binary to trinary. The robustness of the proposed algorithm is validated on Middlebury datasets, synthetic data, and real world data captured by a low-light camera in darkness. The results show that the proposed algorithm has better performance and higher matching rate among top-ranked algorithms on low signal-to-noise ratio data and high accuracy on the Middlebury benchmark datasets.
The influence of adhesive bonding and curing on the accuracy of mirror surface shape was analyzed to realize low-stress assembly of large aperture mirror. Firstly, based on Hooke's law, a curing shrinkage stress equation was deduced, taking deformation of the mirror and support structure into account under the boundary condition of continuous edge bond, and key parameters effecting mirror deformation were obtained. Secondly, for a 514mm ULE spectrometer primary mirror with an inserts structure mosaiced and bonded on mirror-back, an equivalent linear expansion coefficient method was used for finite element modeling. The shrinkage stress at the bond edge of mirror and the mirror surface shape were analyzed. It’s found that adhesive shrinkage has a significant effect on the mirror surface shape. Finally, the inserts structure of mirror assembly was optimized. In contrast to the non-optimum structure, the average stress of adhesive surface caused by adhesive curing shrinkage reduced from 0.28MPa to 0.18MPa, and the mirror surface shape (Root Mean Square, RMS) reduced from 0.029λ to 0.017λ. Finite element analysis results of the mirror assembly were given at last, surface shape accuracy (RMS) of mirror is 0.012λ under a load case of 1g gravity, and the first-order natural frequency of the component is 216 Hz. The obtained results showed that a suitable optimized support structure can effectively relieve adhesive curing stress, and also satisfy the design requirements for both the static and dynamic stiffness.
Large aperture static interferometer spectrometer (LASIS) use the method of push-boom to get the geometric and spectral characteristics of ground target, the particularity of principle requires the movement of satellite must be in the same direction with spectrometers detectors. Drift angle of satellite leading to abnormal image shifts in the column direction which should be perpendicular to the detector and can seriously affect the spectrum recovery precision of collected data. This paper analyzes the influence mechanism of drift angle for spectrum recovery precision. Simulation based on the actual on-orbit data analyses the effects of different drift angle of relative mean deviation and relative secondary deviation rehabilitation of the spectrum, besides the influence of spectral angle similarity. These studies have shown that, when the lateral deviation due to the drift angle on the across track is less than 0.3 pixel, the effect for the relative mean deviation of the inversive spectra will be no more than 7%. when the lateral deviation due to the drift angle on the across track is larger than one pixel, even though the resampling correction is proceeded, the restored spectral data cube still shows an relative mean error more than 10%, which seriously affect the availability of spectral data.
Trajectory prediction is essential for the maneuvering target tracking. Nowadays, one of the major challenges for precise prediction of position and velocity of one maneuvering target is the mismatching of target motion model and the movement mode the target performs. In order to solve this problem, the interacting multiple model method is proposed, which is able to adopt the current model to match the target motion mode so that the precision of prediction can be improved. One of the major problems of the interacting mult iple model methods is the selection of models for the algorithm. In this paper, such three models as constant velocity model, exponential increasing accelerat ion model, and the generalized coordinated turn model is selected. Afterwards, one simulat ion to verify the validation of the algorithm is performed, and it indicates that the interacting mult iple model methods with the specific models utilized in this paper does have the ability to track maneuvering target quite precisely.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.