Acta Geodaetica et Cartographica Sinica

Data logic structure and key technologies on intelligent high-precision map

LIU Jingnan, ZHAN Jiao, GUO Chi, LI Ying, WU Hangbin, HUANG He

2019, 48(8): 939-953. doi:10.11947/j.AGCS.2019.20190125

Asbtract ( )

HTML

PDF (7732KB) ( )

References | Related Articles | Metrics

This paper takes autonomous driving and driverless as the research object, discusses and defines intelligent high-precision map. In this paper, intelligent high-precision map is considered as a key link of future travel, and a carrier of real-time perception of traffic resources in the entire space-time range, and the criterion for the operation and control of the whole process of the vehicle. As a new form of map, intelligent high-precision map has distinctive features in terms of cartography theory and application requirements compared with traditional navigation electronic maps. In order to promote the development of research and application of intelligent high-precision map, it is necessary to analyze and discuss its key features and problems. This paper proposes the information transmission model from the cartography theory; combines the wheeled robot's control flow from practical application, proposes the data logic structure of intelligent high-precision map, and analyzes its application in autonomous driving; summarizes the computing mode of "crowdsourcing+edge-cloud collaborative computing", and carries out key technical analysis on how to improve the quality of crowdsourced data; analyzes the effective application scenarios of intelligent high-precision map in the future; Finally, presents some thoughts and suggestions for the future development of this field.

Probabilities of two types of errors in sampling inspection for surveying and mapping products

CAI Yanhui, CHENG Pengfei, ZHANG Li, XU Yantian

2019, 48(8): 954-959. doi:10.11947/j.AGCS.2019.20180108

Asbtract ( )

HTML

PDF (896KB) ( )

References | Related Articles | Metrics

Based on theory of hypothesis testing, a function of calculating probabilities of two types of errors is proposed, which is adopted inanalyzing the advantages and disadvantages of different sampling schemes given in the National Standard GB/T24356-2009. Under the circumstance of large lot size, the sampling inspection should be carried out lot by lot with the lot size less than 200 and the least lot number. Investigations on such procedure are made base on the probability distributions of two types of errors. Numerical results of type I error show that the whole lot inspection is equivalent to the sub-lot inspections sequence from the point of probability of positive null hypothesis. At the same time, experiment results of type Ⅱ error with the typical sampling scheme show that the conclusions of sampling inspection according to the requests of GB/T24356-2009 may be badly wrong under the situation of low percentage of defectiveness.

Solid lanes extraction from mobile laser scanning point clouds

FANG Lina, HUANG Zhiwen, LUO Haifeng, CHEN Chongcheng

2019, 48(8): 960-974. doi:10.11947/j.AGCS.2019.20180579

Asbtract ( )

HTML

PDF (15282KB) ( )

References | Related Articles | Metrics

This paper presented a novel method for solid lanes extraction from Mobile laser scanning (MLS) point clouds. The proposed method firstly removed the off-ground point clouds and then calculated the scanning distance between the points of road surface and sensors. Then, the reflective intensity data of road surface were transformed into relative values to overcome the influence of the scanning distance, the points' density, abrasion and roughness of road surface block by block. After the intensity enhancement, road markings were separated from the road surface based on the k-means clustering and connected component. In order to deal with the problem of under-segmentation and over-segmentation caused by the adhesion of solid lines and stop lines or other entrance markings, some features of geometric shape and the spatial distribution were then used to refine the results of intensity segmentation by the Normalized Cuts. Finally, the semantic structure information of road markings was explored to separate the solid lines from other road markings like zebra crossings, dashed lines. Experiments were undertaken to evaluate the validities of the proposed method with four test data sets acquired from different MLS systems. Quantitative evaluations on four MLS data sets indicated that the proposed method achieved a Precision, Recall and F₁-Measure of 95.98%, 91.87% and 95.55%, respectively, which validated that the proposed method has achieved promising performance.

Naive Bayes-based automatic classification method of tree-like river network supported by cases

DUAN Peixiang, QIAN Haizhong, HE Haiwei, XIE Limin, LUO Denghan

2019, 48(8): 975-984. doi:10.11947/j.AGCS.2019.20180370

Asbtract ( )

HTML

PDF (4514KB) ( )

References | Related Articles | Metrics

River classification is the key to the generalization of tree-like river network. Most of the existing methods mainly identify the main and tributary according to the local geometric characteristics of the reach, and less consider the overall structural characteristics of the river and river network. The weight setting in the use of multi-index comprehensive evaluation lacks of scientific methods, with less utilization of generalization knowledge, and the flexibility of the application needs to be improved. Focusing on these problems, from the perspective of case-based studying, this paper proposes an automatic classification method of tree-like river network based on naive bayes for the identification of main and tributary of reaches. Firstly, the case of the main tributary classification is extracted from the existing data, and the main-tributary classification model is trained by using the naive bayes method. For the new tree-like river network to be classified, starting from the estuary, from the downstream to the upstream the classification model is used to calculate the probability that each upstream section in the intersection is classified as the mainstream. The upstream section with the highest probability is taken as the mainstream section, and the mainstream sections are connected to the mainstream rivers in turn. The above steps are repeated for the tributaries to carry out the hierarchical structuring process to achieve river classification. The experiment proves that this method can imitate the expert's intention well, and the main and tributary of the tree-like river network are well identified, and a reasonable hierarchical structure is constructed. The classification effect is good.

Salient feature extraction method for hyperspectral image classification

YU Anzhu, LIU Bing, XING Zhipeng, YANG Fan, YANG Qimiao

2019, 48(8): 985-995. doi:10.11947/j.AGCS.2019.20180499

Asbtract ( )

HTML

PDF (10862KB) ( )

References | Related Articles | Metrics

Aiming at the problem of hyperspectral image classification, a salient feature extraction method is proposed. Firstly, the method uses a superpixel segmentation algorithm to divide three adjacent bands of hyperspectral image into several small regions. Then, the salient features of different regions are calculated based on the small regions. Finally, the sliding window method with a size of 3 steps is used along the spectral direction to obtain the salient features of all bands. The extracted saliency features are further combined with the spectral features, and the combined features are fed into a support vector machine for classification. The classification experiments were carried out on three hyperspectral image datasets including Pavia University, Indian Pines and Salinas. The experimental results show that compared with the traditional spatial feature extraction method and the convolutional neural network based methods, the extracted salient features can obtain higher classification accuracy. Combining salient features and spectral features can further improve classification accuracy.

An improved endmember extraction method of mathematical morphology based on PPI algorithm

XU Jun, WANG Cailing, WANG Li

2019, 48(8): 996-1003. doi:10.11947/j.AGCS.2019.20180475

Asbtract ( )

HTML

PDF (1867KB) ( )

References | Related Articles | Metrics

Automated morphological endmember extraction(AMEE) algorithm defines the spectral angular distance between the purest pixel and the most mixed pixel in the structural element as the morphological eccentricity index(MEI) to quantitatively denote the purity of the pixel. However, the most mixed pixels as the reference standard are not the same in different structural elements, especially when the pure pixels account for the majority of the structural elements, the mean spectrum of all the pixels will be closer to the pure pixels. At this time, the higher the MEI, the lower the purity of the pixel. To solve this problem, a novel endmember extraction algorithm is proposed in this paper which combines the pixel purity index (PPI) algorithm with AMEE algorithm and is named PPI-AMEE. In the structural element, the PPI is used to replace the MEI index in the AMEE algorithm to find the purest pixel. When the structural element is transformed, only the purest pixel can always be projected to the two ends of the randomly generated line, therefore the PPI value of the purest pixel will increase continuously, while the PPI value of the other pixels will not increase continuously. The PPI value of each pixel is accumulated and recorded until the iterative termination condition is satisfied, and a PPI image is finally obtained. The endmembers are selected from the pixels with higher PPI value. The PPI-AMEE algorithm runs the PPI algorithm in relatively small structural elements, and then processes the whole image with the expansion operation of mathematical morphology, which takes into account both the spectral and spatial information of the image. In the experiment, AVIRIS hyperspectral data from Cuprite area, Nevada, USA are used to validate the proposed PPI-AMEE algorithm. The experimental results show that the endmember extraction accuracy of PPI-AMEE algorithm is better than that of AMEE algorithm and PPI algorithm on the whole.

Complex urban boundary extraction based on topological interpolation and spectral feature

YU Linan, NING Xiaogang, WANG Hao, LIU Jiping

2019, 48(8): 1004-1013. doi:10.11947/j.AGCS.2019.20180126

Asbtract ( )

HTML

PDF (17613KB) ( )

References | Related Articles | Metrics

Complex urban boundary usually has buildings with greatly different size, rugged forest area and scattered small buildings, which makes many urban extraction algorithms ineffective. In order to solve this problem, a method based on topological interpolation theory and spectral featureis proposed, and it can be applied to complex urban boundary extraction from high resolution imagery. On the one hand, feature corners of big buildings are inserted to avoid the urban boundary lose caused by the sparse feature points. On the other hand, spectral feature is made full use of to remove the redundancy feature points brought by rugged forest area and scattered small buildings, which can reduce the wrong boundary extraction. The experiments and compared analysis show that the new method has a good robustness and high accuracy.

Dimensionality reduction method for hyperspectral images based on weighted spatial-spectral combined preserving embedding

HUANG Hong, SHI Guangyao, DUAN Yule, ZHANG Limei

2019, 48(8): 1014-1024. doi:10.11947/j.AGCS.2019.20180229

Asbtract ( )

HTML

PDF (3863KB) ( )

References | Related Articles | Metrics

Hyperspectral image (HSI) contains a large number of spectral bands, which easily leads to the curse of dimensionality. However, the traditional manifold learning methods generally only consider the spectral features, while the spatial information of HSI is ignored. To overcome this shortcoming, it is proposed that an unsupervised dimensionality reduction algorithm called weighted spatial-spectral combined preserving embedding (WSCPE) for HSI classification. Firstly, the proposed algorithm uses a weighted mean filter (WMF) to filter the image, which can reduce the influence of background noise. Then, according to the spatial consistency property of HSI, it adopts the weighted spatial-spectral combined distance (WSCD) to fuse the spectral and spatial information of pixels to effectively select the spatial-spectral neighbors of each pixel. Finally, the proposed method explores the coordinate distances between pixels and their spatial-spectral neighbors to perform manifold reconstruction, and the low-dimensional discriminative features are extracted for HSI classification. The experimental results on PaviaU and Indian Pines datasets indicate that the overall classification accuracies of the proposed method reached 98.89% and 95.47%, respectively. The WSCPE method not only discovers the intrinsic manifold structure of HSI data, but also effectively integrates the spatial-spectral combined information, which enhances the classification performance.

Stripe noise removal of remote image based on wavelet variational method

WANG Chang, ZHANG Yongsheng, WANG Xu, JI Song

2019, 48(8): 1025-1037. doi:10.11947/j.AGCS.2019.20180394

Asbtract ( )

HTML

PDF (4096KB) ( )

References | Related Articles | Metrics

In order to avoid the loss of image details in the process of strip noise removal, a method based on wavelet variational method was proposed to remove strip noise of remote images. First, remote image with stripe noise was decomposed by wavelet technology. Second, a stripe preserve variation model was constructed, this model could effectively remove image details from the wavelet horizontal direction high-frequency components in lower layers and only preserve the stripe noise, and the details are effectively separated; a destriping variation model was constructed, this model could effectively preserve the image details while removing the strip noise form the wavelet horizontal direction high-frequency components in the top layers. Finally, the destriping image was obtained by wavelet reconstruction. Experimental results show that the proposed method not only can effectively restrain the stripe noise of remote image, and can be also preserve the image details very well. The quality and contrast of destriping image are the best.

Polarimetric radar image despeckling by iteratively refined nonlocal means

MA Xiaoshuang, WU Penghai

2019, 48(8): 1038-1045. doi:10.11947/j.AGCS.2019.20180034

Asbtract ( )

HTML

PDF (5121KB) ( )

References | Related Articles | Metrics

The presence of speckle degrades the quality of the polarimetric synthetic aperture radar (PolSAR) image, hence despeckling is an essential procedure before using SAR images to obtain land-cover information in most cases. In this paper, a PolSAR filtering method based on iteratively refined nonlocal means is presented. In each iteration step of the proposed method, by considering both the statistical trait of the original image and the information of the image obtained in last iteration, the polarimetric similarity between pixels is refined, so as to improve the estimation results. Experiments on a simulated PolSAR image and two real PolSAR images revealed the positive despeckling performances of our proposed method:the speckle is reduced to a large degree, and the image details, such as the edges and the polarimetric traits, are effectively preserved.

Aircraft detection in remote sensing images using cascade convolutional neural networks

YU Donghang, GUO Haitao, ZHANG Baoming, ZHAO Chuan, LU Jun

2019, 48(8): 1046-1058. doi:10.11947/j.AGCS.2019.20180471

Asbtract ( )

HTML

PDF (27376KB) ( )

References | Related Articles | Metrics

Traditional aircraft detection algorithms which adopt handcraft features have poor performance in complex scene images and recognizing multi-scale objects. Methods using deep convolutional neural networks still face difficulty in dim small target search and recognition in large images with complex background. Aiming at these problems, a coarse-to-fine algorithm for aircraft detection in remote sensing images using cascade convolutional neural networks is proposed. To quickly and effectively acquire suspicious regions of interest (ROI), the whole image is searched by a small and shallow fully convolutional neural network which could deal with images of any size. Then deeper convolutional neural networks are used to refine the classification and location of the ROIs. A multilayer perceptron is introduced to the convolutional layer to improve identification capability of the convolutional neural networks and the strategies of multi-task learning and offline hard example mining are adopted in the process of training. At the detecting stage, the image pyramid is constructed and the redundant windows could be eliminated by the non-maximal suppression. Multiple datasets are tested and the results show that the proposed method has higher accuracy and stronger robustness and provides a fast and efficient solution for object detection in large remote sensing images.

Theoretical analysis of soil freeze/thaw process on DDM waveform and multipath in order for GNSS-R/IR applications

WU Xuerui, XIA Junming, BAI Weihua, ZHANG Xinggang

2019, 48(8): 1059-1066. doi:10.11947/j.AGCS.2019.20180038

Asbtract ( )

HTML

PDF (3162KB) ( )

References | Related Articles | Metrics

In order to extend the applications GNSS-R/IR(GNSS-reflectometry/interferometric reflectometry) remote sensing technique to soil freeze/thaw process detection, soil(frozen/thawn) mixing permittivity models are employed to calculate the soil permittivity. Bistatic full-polarization coherent scattering model and random roughness surface scattering model are used to calculate the coherent and non-coherent scattering, which result in the variations of GPS multipath observables and DDM(delay Doppler map) waveforms, respectively. When the soil freeze/thaw process occurs, theoretical simulations indicate that soil freeze/thaw process induce the abrupt permittivity changes and affect the obvious variations of GNSS-R/IR signals. In this way, theoretical fundamentals for soil freeze/thaw process detections are presented.

Public DEM-aided geopositioning without ground control points for chinese satellite imagery

CHEN Xiaowei

2019, 48(8): 1067-1067. doi:10.11947/j.AGCS.2019.20180354

Asbtract ( )

HTML

PDF (662KB) ( )

Related Articles | Metrics

Precise orbit determination and the earth gravity field recovery by acceleration approach for Swarm

ZHANG Bingbing

2019, 48(8): 1068-1068. doi:10.11947/j.AGCS.2019.20180377

Asbtract ( )

HTML

PDF (670KB) ( )

Related Articles | Metrics

Statistical approaches for mining geospatial association patterns

HE Zhanjun

2019, 48(8): 1069-1069. doi:10.11947/j.AGCS.2019.20180411

Asbtract ( )

HTML

PDF (646KB) ( )

Related Articles | Metrics

Study on detection methods of bubble plumes based on multibeam water column data

MENG Junxia

2019, 48(8): 1070-1070. doi:10.11947/j.AGCS.2019.20180430

Asbtract ( )

HTML

PDF (665KB) ( )

Related Articles | Metrics

Lake water storage changes over the Qinghai-Tibetan Plateau from multi-mission satellite data and its influencing factors analysis

HUANG Zhengkai

2019, 48(8): 1071-1071. doi:10.11947/j.AGCS.2019.20180428

Asbtract ( )

HTML

PDF (649KB) ( )

Related Articles | Metrics

Research on intelligent clustering learning algorithm for GNSS data

ZHOU Xiangbing

2019, 48(8): 1072-1072. doi:10.11947/j.AGCS.2019.20180448

Asbtract ( )

HTML

PDF (687KB) ( )

Related Articles | Metrics

Table of Content