Joint Processing of Uav Imagery and Terrestrial Mobile Mapping System Data for Very High Resolution City Modeling

New aerial sensors and platforms (e.g., unmanned aerial vehicles (UAVs)) are capable of providing ultra-high resolution remote sensing data (less than a 30-cm ground sampling distance (GSD)). This type of data is an important source for interpreting sub-building level objects; however, it has not yet been explored. The large-scale differences of urban objects, the high spectral variability and the large perspective effect bring difficulties to the design of descriptive features. Therefore, features representing the spatial information of the objects are essential for dealing with the spectral ambiguity. In this paper, we proposed a dual morphology top-hat profile (DMTHP) using both morphology reconstruction and erosion with different granularities. Due to the high dimensional feature space, we have proposed an adaptive scale selection procedure to reduce the feature dimension according to the training samples. The DMTHP is extracted from both images and Digital Surface Models (DSM) to obtain complimentary information. The random forest classifier is used to classify the features hierarchically. Quantitative experimental results on aerial images with 9-cm and UAV images with 5-cm GSD are performed. Under our experiments, improvements of 10% and 2% in overall accuracy are obtained in comparison with the well-known differential morphological profile (DMP) feature, and superior performance is observed over other tested features. Large format data with 20,000 × 20,000 pixels are used to perform a qualitative experiment using the proposed method, which shows its promising potential. The experiments also demonstrate that the DSM information has greatly enhanced the classification accuracy. In the best case in our experiment, it gives rise to a classification accuracy from 63.93% (spectral information only) to 94.48% (the proposed method).

“…The orthophoto and DSM are generated from a UAV mission, as described in [9,40], with a 5-cm GSD. The DSM is generated using a hierarchical semi-global matching algorithm [41,42].…”

Section: Experiments With Test Datasetmentioning

confidence: 99%

Classification of Ultra-High Resolution Orthophotos Combined with DSM Using a Dual Morphological Top Hat Profile

Zhang

Qin²,

Huang

et al. 2015

“…With 16 mm focal length and 5 micron pixel size, the flying height is 150 meters above the ground, producing images with roughly 5 cm GSD. The interior and exterior orientations of image in date 1 were computed with GCPs from land surveys from one of our previous 3D modeling project [48], with RMSE of 0.08, 0.06, 0.06 in X, Y, Z dimension for check points. The co-registration of these two dates is illustrated in the "Data Acquisition and Pre-processing" section, with average image residual of 0.6 pixels.…”

Section: Experiments and Discussionmentioning

confidence: 99%

An Object-Based Hierarchical Method for Change Detection Using Unmanned Aerial Vehicle Images

Qin¹

2014

There have been increasing demands for automatically monitoring urban areas in very high detail, and the Unmanned Aerial Vehicle (UAV) with auto-navigation (AUNA) system offers such capability. This study proposes an object-based hierarchical method to detect changes from UAV images taken at different times. It consists of several steps. In the first step, an octocopter with AUNA capability is used to acquire images at different dates. These images are registered automatically, based on SIFT (Scale-Invariant Feature Transform) feature points, via the general bundle adjustment framework. Thus, the Digital Surface Models (DSMs) and orthophotos can be generated for raster-based change analysis. In the next step, a multi-primitive segmentation method combining the spectral and geometric information is proposed for object-based analysis. In the final step, a multi-criteria decision analysis is carried out concerning the height, spectral and geometric coherence, and shape regularity for change determination. Experiments based on UAV images with five-centimeter ground resolution demonstrate the effectiveness of the proposed method, leading to the conclusion that this method is practically applicable for frequent monitoring.

“…This supposes a major motivation for tandem and simultaneous aerial-ground mapping concepts, as we propose in mapKITE. We identify initiatives addressing the fusion of UAS images and terrestrial ones [2,3] and the combination of UAS photogrammetry-based 3D models with terrestrial laser scanner point clouds [4]-however, to the best of our knowledge, there is no available approach for simultaneous and kinematic geodata acquisition mode.…”

Section: Introductionmentioning

confidence: 99%

First Results of a Tandem Terrestrial-Unmanned Aerial mapKITE System with Kinematic Ground Control Points for Corridor Mapping

Molina¹,

Blázquez²,

Cucci

et al. 2017

Abstract:In this article, we report about the first results of the mapKITE system, a tandem terrestrial-aerial concept for geodata acquisition and processing, obtained in corridor mapping missions. The system combines an Unmanned Aerial System (UAS) and a Terrestrial Mobile Mapping System (TMMS) operated in a singular way: real-time waypoints are computed from the TMMS platform and sent to the UAS in a follow-me scheme. This approach leads to a simultaneous acquisition of aerial-plus-ground geodata and, moreover, opens the door to an advanced post-processing approach for sensor orientation. The current contribution focuses on analysing the impact of the new, dynamic Kinematic Ground Control Points (KGCPs), which arise inherently from the mapKITE paradigm, as an alternative to conventional, costly Ground Control Points (GCPs). In the frame of a mapKITE campaign carried out in June 2016, we present results entailing sensor orientation and calibration accuracy assessment through ground check points, and precision and correlation analysis of self-calibration parameters' estimation. Conclusions indicate that the mapKITE concept eliminates the need for GCPs when using only KGCPs plus a couple of GCPs at each corridor end, achieving check point horizontal accuracy of µ E,N ≈ 1.7 px (3.4 cm) and µ h ≈ 4.3 px (8.6 cm). Since obtained from a simplified version of the system, these preliminary results are encouraging from a future perspective.