2017 Vol. 42, No. 11
We introduce the background of civil-military integration strategy, analyze the achievements and challenges of the present situation of China's space-based information system, and discuss the strategic significance of the construction of a civil military integrated space-based real-time information service system. A proposal for the construction of communication-navigation-remote sensing integrated space-based information service system (including positioning, navigation, timing, remote sensing, and communication functions, PNTRC) is put forward. The ultimate goal is a implementation strategy, technical roadmap, and prospective PNTRC application. Key technologies for the construction of PNTRC are analyzed, some suggestions for key technologies are presented.
The sparks of an intelligent revolution, triggered by artificial intelligence, are spreading to all walks of life. The challenges and opportunities for methods, technology, as well as industrial and business models of mapping, surveying, and location based services, are a question that has to be considered in depth. The development paths and trends in artificial intelligence, are discussed for an understanding of the characteristics of the new generation of artificial intelligence. The opportunities and challenges brought by the coming age of intelligence are analyzed. Focusing on the field of mapping, surveying and location based services, this paper argues that this industry will not disappear but has a new impetus towards transformation. The possible paths for transformation are discussed. As an academic discipline, it will not expand, but fuse and cross disciplinary borders. As a career, it will make blue collar work an evaporating endeavor, while a number of makers and thinkers will become leading figures. The mapping, surveying, and location based service industry needs to fully tap their own advantages, with systemic thinking, space-time thinking and creative thinking as the key to open the door of the age of intelligence. Thereby, an overall transformation for mapping, surveying and location based service industry to be green, intelligent, pervasive development can be achieved.
We outline the history of photogrammetry from the aspects of, perspective geometry, camera, platform, measure methods and measure instruments, and summarize previous contributions to photogrammetry. A brief review of computer vision history is given. The tight connections between computer vision and photogrammetry are discussed in terms of geometric principles, and some differences in applications are also considered. From the aspect of semantics, we analyze the development of remote sensing and its relations to machine learning and computer vision, including their common approaches and different applications. The prevailing deep learning raised from connectionism is also reviewed and its successful applications in photogrammetry are analyzed. At last, we expect that the future development of photogrammetry will be more tightly cross-integrated with computer vision, machine learning and artificial intelligence.
An Unmanned Aerial Vehicle (UAV) is reusable, consisting of power system and unmanned autopilot controller; while a Unmanned Aircraft System (UAS) is a system controlled manually, automatically or independently to perform different kinds of tasks This article summarizes and analyzes the characteristics of UAV and UAS in different historical stages; secondly, this article emphasizes UAV and UAS requirements for Artificial Intelligence technologies; finally, this article discusses potential influences on UAS. A UAS usually consists of UAV platforms, payloads for tasks, datalink devices, information processing devices, and integrated support equipment. Research on UAS and UAV started from the beginning of the 20th century, and with the development of electronics, mechanics, material science, and computer science; UAS has rapidly developed over the last century, especially the latest thirty or forty years. With the rapid development of Artificial Intelligence over the first 20 years of the 21st century ushered in a new stage of UAS and UAV development.
Space-borne platforms are indispensable for sea target surveillance. Multi-source information fusion can improve the accuracy of detection, confidence of recognition, coverage of spatial-temporal and reduce the fuzziness of inference. To improve the capability of maritime target wide-area searches, accurate identification, persistent surveillance and operational response, both space-based maritime target surveillance systems and multi satellite information fusion need to be further studied. The characteristics of maritime environment and maritime target for maritime target information awareness and fusion are reviewed and the developments, data characteristics and disadvantages of current space-based maritime surveillance systems are summarized. The key technologies and future development of space-based maritime surveillance are discussed in the fields of satellites networking, temporal-spatial fusion and information fusion.
A multi-sensor combined imaging system is proposed based on the aerial survey requirements for national level emergencies. This system integrates digital cameras, video and infrared cameras, LiDAR and MiniSAR into a unified benchmark of space and time. It can perform optical large format imaging, daytime and night video imaging for real time transfer, 3D laser point clouds acquisition, and all-weather microwave sensing respectively and simultaneously on a flying platform. The advantages of combined wide-angle cameras for improving image quality and operation efficiency are described in detail. Based on this design, an experimental system was installed in a medium duration UAV, to make test flights. Observed daytime and night data are shown.
Efficient and automatic road pavement defect detection technology has become increasingly important given the transition of China's highway development from the large-scale construction stage to the large-scale conservation stage. We briefly review the study of road surface defect detection over the past two decades, and present a new method for road surface defect detection using 3D laser profiling techniques. Road surface defect detection involves the use of optical sensors such as CCD cameras and laser scanners to collect the information of road surface, and pattern recognition and machine learning algorithms to automatically locate the defects and identify the categories of defects. Comparative experiments in the laboratory and test practices on the real roads show that the road surface defect detection technology using 3D laser profiling has stronger environmental adaptability and higher recognition precision than the traditional technology based on 2D visible light imaging.
3D power line reconstruction is one motivation for power line patrols using LiDAR systems mounted on helicopters. Two new models for 3D power line reconstruction are proposed. The first model combines a line segment and a catenary, while the second model is a combination of a line segment and a parabola. In both models, the line segment is located in the xy-plane, and the catenary or the parabola is on a vertical plane through the line segment. The two proposed models are innovative in that a scalar factor was employed as the parameter for the catenary and the parabola. The scalar factor is calculated by projecting a 2D point on the xy-plane onto a fitted line by the 2D points. Specifically, a 2D point is constructed by projecting a 3D LiDAR point on the xy-plane. Six experimental data sets and four criterion were used to test the six methods for 3D power line reconstruction. These experiments suggest that he second model has the highest efficiency and reconstruction accuracy. Moreover, three factors, including the selection of the vertical plane, selection of the projected models in the vertical plane, and consideration of the errors, have significant influences on the performance of the methods for 3D power line reconstructions.
We present the principals for BDS and BDS/GPS real-time kinematic (RTK) positioning. A experiment was conducted on Songshan calibration field using a UAV-based low altitude photograph platform which carried a double frequency GNSS system simultaneously receiving BDS and GPS signals. By comparing the positioning result of GPS, BDS, and the BDS/GPS-combined RTK models, and using these three groups of camera exposure station coordinates to carry out GNSS-supported aerotriangulation. These experiments demonstrate that the performance of GPS, BDS and BDS/GPS-combined RTK positioning model are very close in bundle block adjustment. For all three groups, the overall root mean square error was all lower than±0.17 m for the planimetry and±0.25 m for height, which satisfies the Chinese National specifications of aerial photogrammetric 1:500 mapping in hilly regions.
On-orbit geometric calibration of remote sensing satellites is essential to the geometric quality of its imagery. This paper reviews the current status and progress in optical satellite on-orbit geometric calibration technology, based on ground calibration site and space resection theory. The construction status of calibration sites and calibration results of SPOT, IKONOS, GeoEye, ZY-3 and other mainstream high-resolution optical satellites are summarized. problems existing in current calibration methods caused by dependence on reference images of a calibration site are analyzed considering improvements in the resolution of optical satellite imagery and timeliness requirements of on-orbit calibration, For the demand of camera calibration without the constraint of ground test site, the existing work in the field of computer vision and block adjustment are discussed. The merits and faults of each self-calibration method are compared, and a self-calibration method for optical satellites based on multiple-angle imaging is introduced, and several self-calibration modes are proposed.
The location accuracy of space-borne laser altimeters suffers from installation errors. Calibration could improve the positioning and height accuracy effectively based on the operation data of satellite in orbit. A novel calibration method is put forward for ZiYuan-3 (02) satellite (ZY3-02) laser altimeter that without waveform. In this approach, the error in the laser distance and positioning calibration is indicated by the terrain matching method. A test area with an exposed surface constrains the orbit laser ranging using SRTM-DEM data, and the initial calibration of the direction of the starter is obtained by approximating the terrain fluctuation trend line. The test results were verified; the relative pointing precision of different orbits data was 20 meters. A series of laser reflectors operated on the ground determine the surface radiometric intensity of the laser footprint. The calibration parameters compensate the laser positioning model; the height accuracy of the ZY3-02 satellite laser altimeter was 0.5 meter on a flat surface.
This paper proposes a multi-resolution TIN surface modeling and visualization method for coastal topography in a virtual globe system. a tile-pyramid model is used to organize the multi-resolution TIN data to solve the problems stemming from the storage of large-scale reference TIN data. A greedy insertion algorithm was used to construct the multi-resolution TIN pyramid. In order to efficiently index the global TIN data, we also propose a novel virtual node (VN) model to guarantee the seamless partition of the multi-resolution TIN triangulation into tiles and storage in a chunked TIN pyramid. Aiming to avoid terrain cracks, we also developed an encoding and decoding approach to match triangulation of different resolutions. Our method was implemented in a virtual globe system. The results show that, the method can dramatically reduce the amount of triangles for rendering the ocean surface, while achieving high fidelity in representation of the islands and coastal lands. This approach overcomes deficiencies of the traditional Grid-based virtual globes.
We put forward a method for image classification based on data gravitation. The quality of the data particles we use is the feature of the images(such as the fractal dimension of the image). For each kind of training data particle set, we use the mean of several image characteristics mi in the set to be the image characteristic, and the number of images wi to be its weight. So the quality of the i-thtraining data particle set is wi mi, and the data particle for inspection is atomic data particle with the mass of 1. Assuming that the characteristic of the image data particle for inspection j is tmj, the distance between the i-th training data particle set and the data particles for inspection j of the image is |mi-tmj|. Assuming that there are three different catagories of images, we choose the part of images from all kinds in order to compose three kinds of training data particle set, then work out the characteristic mean of data particle set of each kind and characteristic value of a data particle for inspection, which can be used to calculate the gravitation between data particle set of each kind and characteristic value of data particle for inspection. The kind with largest gravitation is the one of data particle for inspection. It is proved by the experimental result that the method for image classification based on data gravitation has a certain advantage.
Unmanned aerial vehicle (UAV) scheduling in emergency surveying and mapping (ESM) has complicated space and time constraints, with diverse scheduling schemes. The existing scheduling methods are generally based on human experience, and not reliable. In order to realize the rapid response of post disaster ESM, a fast scheduling method for UAV resources is proposed. The method comprehensively considers the constraints of the mission requirements, priorities, time windows, the geographical environment of the working area, and the ability of the UAV. A UAV scheduling model for ESM, is proposed that makes mission profit efficiency, mission completion rate, and scheduling risk as optimization objectives. This solution is based on ant colony algorithm. Experimental results verify the effectiveness of the scheduling method.
With diversification of data acquisitions and associated technologies, there is steady accumulation of multi-source spatial data, prompting more research on effective and fast data integration to provide information and technical support for geospatial research and applications. In this paper, procedures for integrating multi-source data are referred to as data fusion. Research developments are reviewed by considering them as classic and extended types of methods, respectively. To review the progress in classic methods for data fusion and related data processing, this paper clarifies some of the relatively well-established methods using a typology of object-and field-based models of spatial data. For extended methods, the paper discusses multi-point geostatistics, statistical-mechanistic methods, multi-scale signal analysis and reconstruction, and information-theoretic strategies. Some of the issues, such as scale mismatch, semantic inconsistency, and the temporal dimension, are discussed in the hope of better supporting the fusion of multi-sourced and heterogeneous data. Further research will enhance theoretical foundations of geographic information to enrich methodologies for spatial data and their analyses, and help add to the applicability and value of spatial information.
In map generalization, spatial conflicts may occur either between different sections of one single road or between two different roads. Displacement is required when such conflicts emerge. At present, the energy minimization approach is the dominant method for understanding road displacement; but model parameter control and displacement propagation need to be studied further in order to control road displacement effectively. The energy minimization model is improved by setting up obstacle points, calculating distance of displacement propagation, and adjusting model parameters automatically. These improvement measures are all guided by cartographic rules. The effectiveness and availability of this proposed method are verified by experiments.
The realization and maintenance of the China geodetic coordinate system 2000 (CGCS2000) depends on GPS technologies, and therefore is a national security risk. BDS began offering services to customers in the Asia-Pacific region in December 2012. It provides a good opportunity to realize and maintain the national terrestrial reference frame (TRF) using BDS system. We focus on how to use BDS observations of BDS/GPS stations with dual-mode receivers to realize the National TRF, aligned to the international terrestrial reference frame (ITRF). Results show that the precision of the horizontal velocity derived from BDS data are comparable to those derived from GPS data at about 2-3 mm/a, using more than two years of observation data. The RMS of the station coordinate residuals, with the linear velocity removed, are better than 0.8 cm and 1.7 cm for the horizontals and verticals, respectively. The current precision of BDS data can allow observation of seasonal displacements in the verticals. System errors related to the longitude and latitude are found in the coordinate differences between BDS and GPS data based on PPP mode. Thus, the current BDS system can be used to realize and maintain the national terrestrial reference frame at cm level.
A seamless coastal gravimetric quasi-geoid model was computed using a combination of heterogeneous gravity data. A comparison to the precise GNSS/leveling derived height anomalies, accurately determined the discrepancy of the 1985 national height datum with respect to the gravimetric quasi-geoid. Based on the numerical results, the discrepancy of the 1985 national height datum with respect to the gravimetric quasi-geoid with W0 defined by IERS2010 standard is 30.4±1.0 cm. After transmitting the 1985 national height datum to Hainan Island through the Qiongzhou Strait in 1980s, the regional height datum of Hainan Island showed no significant bias at the 1 cm accuracy level. The standard deviation of the differences between the GNSS/leveling derived height anomalies and the computed quasi-geoid model was 4 cm. The quasi-geoid is seamless across the borders of the 13 provinces and cities. Research findings show that during the last ten years, land subsidence is extensive over most areas of Tianjin except for its northwest corner, where the land subsidence over southeast area reaches up to 7 cm. The land subsidence of Shanghai is about 7 cm, and there are various degrees of land subsidence in some parts of Zhejiang, Jiangsu and Fujian as well.
A time system and reference frame are two fundamental parameters in a satellite navigation system (GNSS), which consists of coordinates and clocks of satellites and stations. A comparison of these precise parameters of IGS Analysis Center shows apparent differences. The differences include constant offsets and periodical signals, where the periodical signals of satellite orbits and satellite clocks are highly correlated. We analyzed the characteristics of the periodical signals and derived their amplitudes and frequencies. The derived periodical terms were applied for the correction of AC satellite orbits and clocks. The numbers of epochs where the agreement between AC precise products increased, amounted to 95.3% and 85.6% of all samples for orbits and clocks, respectively. Furthermore, the error in such signals was analyzed and strategies based on the parameter de-correlation are proposed for the GNSS precise data processing.
The number of signal rays passing through a research area are often limited due to the geometric distribution of satellite constellations and receivers. To solve this issue, an approach is proposed that imposes virtual slant signals on the tomographic model. This approach increases the number of signal rays passing through a research area and. improves accuracy. An experiment tested the feasibility and accuracy of proposed approach by comparing with the results using Global Positioning System and meteorological data for 31 days in May, 2013, provided by Hong Kong Satellite Positioning Reference Station Network and those derived from radiosonde data. The Root Mean Square water vapor profile and error statistics for proposed method are all superior to that of traditional method. The experiment shows that the proposed approach can improve the accuracy of tomographic results.
Based on the distribution of aftershock sequences and the geology of Lushan, using the GPS and level measurement coseismal displacement data near the Lushan Earthquake, we constructed a curved fault model of the 2013 Ms 7.0 Lushan earthquake using the grid search and variance component estimation method. We executed a fault slip distribution inversion based on this curved fault model. The result shows that the fault is mainly a thrust fault with a sinistral trend. The earthquake occurred at 4.56~18.25 km in depth, the maximum slip was 0.70 m at the 12.64 km depth, and the corresponding seismic moment release was 8.77×1018N/m, equivalent to a Mw 6.57 earthquake. Based on a check board test of the curved fault model, we can distinguish an asperity shallower than 13 km in the fault depth, with a minimum area of 8 km×8 km.
the coseismic and the preseismic displacements in the regional reference frame were obtained for the M7.8 earthquake on Nov. 13, 2016 in New Zealand based on time series of coordinates at continuous GPS observation stations, available from http://geodesy.unr.edu, in New Zealand and Australia. horizontal displacement time series vectors at CMBL and KAIK in New Zealand and at USUD and TSKB in Japan are shown. The vector time series show clearly with direct visual effect that the coseismic horizontal displacements were sudden and dramatic, and opposite to the accumulated preseismic horizontal displacements. These represent a rebound or elastic rebound of the accumulated preseismic horizontal displacements. Therefore, the accumulated preseismic horizontal displacements were precursors to the earthquakes. Results produced from the abundant GPS observations are used to discuss the rebound and elastic rebound models. Detailed reanalysis on the results of GPS observations of the M9.0 Tohoku earthquake in Japan shows that one principal strain in the region at and around the epicenter was dominated by compression in the east direction before the earthquake. Moreover, the earthquake and the coseismic displacements were the result of shear rupture at the earthquake faults. This agrees with test of rock failure results, strain analysis, and consistent with earthquake mechanisms.