Mei Zhou

The research of land covers classification based on waveform features correction of full-waveform LiDAR

In order to solve the problem of insufficient classification types and low classification accuracy using traditional discrete LiDAR, in this paper, the waveform features of Full-waveform LiDAR were analyzed and corrected to be used for land covers classification. Firstly, the waveforms were processed, including waveform preprocessing, waveform decomposition and features extraction. The extracted features were distance, amplitude, waveform width and the backscattering cross-section. In order to decrease the differences of features of the same land cover type and further improve the effectiveness of the features for land covers classification, this paper has made comprehensive correction on the extracted features. The features of waveforms obtained in Zhangye were extracted and corrected. It showed that the variance of corrected features can be reduced by about 20% compared to original features. Then classification ability of corrected features was clearly analyzed using the measured waveform data with different characteristics. To further verify whether the corrected features can improve the classification accuracy, this paper has respectively classified typical land covers based on original features and corrected features. Since the features have independently Gaussian distribution, the Gaussian mixture density model (GMDM) was put forward to be the classification model to classify the targets as road, trees, buildings and farmland in this paper. The classification results of these four land cover types were obtained according to the ground truth information gotten from CCD image data of the targets region. It showed that the classification accuracy can be improved by about 8% when the corrected features were used.

Recent development of hyperspectral LiDAR using supercontinuum laser

Hyperspectral Light Detection And Ranging (Hyperspectral LiDAR), a recently developed technique, combines the advantages of the LiDAR and hyperspectral imaging and has been attractive for many applications. Supercontinuum laser (SC laser), a rapidly developing technique offers hyperspectral LiDAR a suitable broadband laser source and makes hyperspectral Lidar become an installation from a theory. In this paper, the recent research and progressing of the hyperspectral LiDAR are reviewed. The hyperspectral LiDAR has been researched in theory, prototype system, instrument, and application experiment. However, the pulse energy of the SC laser is low so that the range of the hyperspectral LiDAR is limited. Moreover, considering the characteristics of sensors and A/D converter, in order to obtain the full waveform of the echo, the repetition rate and the pulse width of the SC laser needs to be limited. Recently, improving the detection ability of hyperspectral LiDAR, especially improving the detection range, is a main research area. A higher energy pulse SC laser, a more sensitive sensor, or some algorithms are applied in hyperspectral LiDAR to improve the detection distance from 12 m to 1.5 km. At present, a lot of research has been focused on this novel technology which would be applied in more applications.

Simultaneous remote sensing image classification and annotation based on the spatial coherent topic model

The traditional LDA models to solve the problem of scene classification lack the spatial relationship between the fragments of images or the parts of targets and linkages between the global and local information, so their performance is usually poor in stability for the images with clutter background. In this paper, a novel method for the simultaneous classification and annotation of remote sensing images with complex scenes is proposed. The Spatially Consistent Topic Model is defined by making full use of the correlation between image classification and annotation. We choose SIFT features, hue features and texture features as the visual words, which help to endow pixels of similar appearance region with the same hidden topic. Competitive results on remote sensing images demonstrate the precision and robustness of the proposed method.

Automatic detection and mapping of urban buildings in high resolution remote sensing images

In high resolution remote sensing images, urban buildings always have characteristics of complex structures and are vulnerable to background interference. For the purpose of detecting and mapping urban buildings automatically in that circumstance, a novel method is proposed in this paper. Firstly, the Conditional Random Field (CRF) is introduced to fuse multiple kinds of features to get the areas objects existing, then we propose a Hierarchical Object Process Model (HOPM), which is used to access to the location of objects as well as accurate depictions of their outline, and finally the corner detection method is utilized to delineate the vector shapes of objects. Competitive results for multiform and complicated urban buildings demonstrate the precision and robustness of the proposed method.

Comparison and research of group refractivity models and atmospheric delay to lidar

Satellite laser range system measures the distance between the satellite and the surface of the earth by figuring out the transit time of laser pulse. The beam is refracted when it goes through the atmosphere. The atmosphere refraction effect causes laser propagation delay and path bending, which is one of the key factors to restrict the accuracy of laser ranging. In order to improve the accuracy of atmospheric refraction delay correction, it is necessary to strengthen the study of atmospheric group refractivity models and atmospheric refraction delay correction method. According to the data of Xuzhou upper air meteorological station, which are the monthly values of upper limit layers for 30 years (1981-2010) in China, three atmospheric group refractivity models were analyzed and compared. The atmospheric refraction delays to LiDAR were calculated by ray tracing method. The differences among the group refractivity models as a function of month or direction angle were given, which lay the foundation for the practical application and precision evaluation of LiDAR.

A Concealed Car Extraction Method Based on Full-Waveform LiDAR Data

Concealed cars extraction from point clouds data acquired by airborne laser scanning has gained its popularity in recent years. However, due to the occlusion effect, the number of laser points for concealed cars under trees is not enough. Thus, the concealed cars extraction is difficult and unreliable. In this paper, 3D point cloud segmentation and classification approach based on full-waveform LiDAR was presented. This approach first employed the autocorrelation coefficient and the echo ratio to determine concealed cars areas. Then the points in the concealed cars areas were segmented with regard to elevation distribution of concealed cars. Based on the previous steps, a strategy integrating backscattered waveform features and the view histogram descriptor was developed to train sample data of concealed cars and generate the feature pattern. Finally concealed cars were classified by pattern matching. The approach was validated by full-waveform LiDAR data and experimental results demonstrated that the presented approach can extract concealed cars with accuracy more than 78.6% in the experiment areas.

Study on road detection method from full-waveform LiDAR data in forested area

Road detection in forested area gained its popularity in recent years, which has an important significance in the field of forest management, fire prevention and so on. The full-waveform LiDAR has unique advantages to forest road extraction, it can record the whole returned waveform while obtaining the information of 3D coordinates of targets points, by processing waveform data, the points density can be increased and the waveform features reflecting the properties of targets can be obtained. This paper deals with road detection in forested area by using full-waveform LiDAR data. Firstly, the returned waveforms were processed, including waveform denoising, waveform decomposition and features extraction. The extracted features were distance, intensity, Full Width at Half Maximum (FVVHM) and back scattering cross section. In order to decrease the differences between the features of same land cover type and further improve the effectiveness of road detection, this paper has made comprehensive correction on the extracted features. Secondly, four pseudo color feature images illustrating the values of extracted features were generated. Then, a region growing algorithm of image segmentation was carried out to retrieve the full roads from the feature images. The algorithm started from a pixel seed which was manually selected. Finally, four different segmentations were merged to obtain the road detection results. It showed that the forest road detection based on full-waveform data could achieve very good effect.

Performance analysis of weak target detection via ground-based synthetic aperture radar

Polarimetric Interferometric Synthetic Aperture Radar (Pol-InSAR) is an emerging technique that combines interferometric SAR and polarimetric SAR techniques and has shown its effectiveness in the detection of buried weak targets. The detection performance is affected by the SAR parameters as well as the covering characteristics. In this paper, the effects of covering characteristics on the detection performance were emphasized and experimentally investigated by a ground-based Pol-InSAR system. Firstly, the detection principle for buried weak target by Pol-InSAR technique was presented, which is based on the use of interferometric coherence variation with polarization. Then the ground-based two dimensional rail (TDR) SAR used for investigation was introduced. Furthermore, the experiment target scene was designed and the effects of different covering type, different covering moisture, and different covering depth on the detection performance of weak targets were shown and analyzed. Preliminary results confirmed the effectiveness of Pol-InSAR technique used for weak target detection and it would be helpful for the further investigation of this technique.