Performance evaluation of classification trees for building detection from aerial images and LiDAR data: a comparison of classification trees models

Salah, Mahmoud; Trinder, John C.; Shaker, Ahmed

doi:10.1080/01431161.2010.507678

Cited by 6 publications

(6 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Discrimination of natural and human-made features using nDSM alone is not Haala and Brenner (1999) suggest that nDSM cannot be used for differentiation and extraction of streets or LULC. However, by combining nDSM with the CHM and intensity surface models, it is possible to effectively highlight built-up areas and enhance discrimination of LULC (Salah et al, 2011).…”

Section: Normalized Digital Surface Modelmentioning

confidence: 99%

LiDAR-Landsat data fusion for large-area assessment of urban land cover: Balancing spatial resolution, data volume and mapping accuracy

Singh

Vogler

Shoemaker

et al. 2012

ISPRS Journal of Photogrammetry and Remote Sensing

115

View full text Add to dashboard Cite

Section: Normalized Digital Surface Modelmentioning

confidence: 99%

LiDAR-Landsat data fusion for large-area assessment of urban land cover: Balancing spatial resolution, data volume and mapping accuracy

Singh

Vogler

Shoemaker

et al. 2012

ISPRS Journal of Photogrammetry and Remote Sensing

115

View full text Add to dashboard Cite

“…Empty cells were assigned elevation values by a nearest neighbor interpolation from neighboring pixels to avoid introducing new elevation values into the generated DTM [53]. The DTM was produced by computing the mean elevation value of the ground returns within each 1.0 mˆ1.0 m grid cell (Figure 2b).…”

Section: Hyperspectral Datamentioning

confidence: 99%

Fusion of Airborne Discrete-Return LiDAR and Hyperspectral Data for Land Cover Classification

Luo

Wang

et al. 2015

Remote Sensing

View full text Add to dashboard Cite

Accurate land cover classification information is a critical variable for many applications. This study presents a method to classify land cover using the fusion data of airborne discrete return LiDAR (Light Detection and Ranging) and CASI (Compact Airborne Spectrographic Imager) hyperspectral data. Four LiDAR-derived images (DTM, DSM, nDSM, and intensity) and CASI data (48 bands) with 1 m spatial resolution were spatially resampled to 2, 4, 8, 10, 20 and 30 m resolutions using the nearest neighbor resampling method. These data were thereafter fused using the layer stacking and principal components analysis (PCA) methods. Land cover was classified by commonly used supervised classifications in remote sensing images, i.e., the support vector machine (SVM) and maximum likelihood (MLC) classifiers. Each classifier was applied to four types of datasets (at seven different spatial resolutions): (1) the layer stacking fusion data; (2) the PCA fusion data; (3) the LiDAR data alone; and (4) the CASI data alone. In this study, the land cover category was classified into seven classes, i.e., buildings, road, water bodies, forests, grassland, cropland and barren land. A total of 56 classification results were produced, and the classification accuracies were assessed and compared. The results show that the classification accuracies produced from two fused datasets were higher than that of the single LiDAR and CASI data at all seven spatial resolutions. Moreover, we find that the layer stacking method produced higher overall classification accuracies than the PCA fusion method using both the SVM and MLC classifiers. The highest classification accuracy obtained (OA = 97.8%, kappa = 0.964) using the SVM classifier on the layer stacking fusion data at 1 m spatial resolution. Compared with the best classification results of the CASI and LiDAR data alone, the overall classification accuracies improved by 9.1% and 19.6%, respectively. Our findings also demonstrated that the SVM classifier generally performed better than the MLC when classifying multisource data; however, none of the classifiers consistently produced higher accuracies at all spatial resolutions.

show abstract

“…The method presents promising results, however, when comparing to the proposed method, it is verified that our method presents greater simplicity in the pre-processing, reached good results, mainly in regions that are confusion among similar spectral classes and regions that suffer shadow influence. Salah et al (2011) used feature attributes generated from RGB aerial image, intensity image, DSM and nDSM in the classification tree procedure to firstly obtain buildings, trees, roads and ground classes. The authors claim that using attributes generated from RGB aerial and intensity images resulted in lower classification accuracies, while using attributes generated from DSM and normalised resulted in significantly higher classification accuracies.…”

Section: Comparison To Other Methodsmentioning

confidence: 99%

“…Alonso and Malpica (2010) integrated ALS elevation data with SPOT 5 medium-resolution (multispectral and panchromatic) images to detect buildings and other objects via Support Vector Machine (SVM) classifiers, resulting in an improvement of 16.78% in the overall accuracy of classification when compared to the results obtained using only image data. Salah et al (2011) used a set of 26 uncorrelated feature attributes derived from original aerial images and ALS data (intensity and height data) in classification trees models for classifying building, trees, roads and ground. The results showed that using attributes generated from the multispectral aerial or intensity images resulted in lower classification accuracies.…”

Section: Introductionmentioning

confidence: 99%

Urban road classification in geometrically integrated high-resolution RGB aerial and laser-derived images using the artificial neural network classification method

Mendes

Póz

2018

International Journal of Image and Data Fusion

View full text Add to dashboard Cite

The problem of automated urban road network extraction is extremely complex because roads in urban scenes strongly interact with other objects. This problem can be simplified if road regions are first isolated using a classification procedure. The isolated road regions can be posteriorly used in tasks of refinement and reconstruction of the road network. This article addresses only the problem of road regions' detection using Artificial Neural Network as classification method. However, in urban areas, the use of spectral data alone commonly leads to the confusion of the road class with other classes in RGB images, such as building roofs and concrete, because these objects may present similar spectral characteristics. To overcome this problem, it is proposed the integration of a high-resolution RGB aerial image with laser-derived images. The classification results showed that the integration of the geometric (height) and radiometric (laser pulse intensity) laser data significantly improved the classification accuracy, also contributing for the better detection of road pixel. The laser intensity data help to overcome the effects of road obstructions caused by shadows and trees. On the other hand, the laser height data help to separate the aboveground objects from those on the ground level.

show abstract

Performance evaluation of classification trees for building detection from aerial images and LiDAR data: a comparison of classification trees models

Cited by 6 publications

References 21 publications

LiDAR-Landsat data fusion for large-area assessment of urban land cover: Balancing spatial resolution, data volume and mapping accuracy

LiDAR-Landsat data fusion for large-area assessment of urban land cover: Balancing spatial resolution, data volume and mapping accuracy

Fusion of Airborne Discrete-Return LiDAR and Hyperspectral Data for Land Cover Classification

Urban road classification in geometrically integrated high-resolution RGB aerial and laser-derived images using the artificial neural network classification method

Contact Info

Product

Resources

About