SM
S.C. Madanu
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
2 records found
1
Accurately classifying laser-scanned point cloud data remains a critical challenge in geospatial analysis, particularly due to the complexity and volume of the data. This thesis presents a novel, confidence-aware deep learning framework designed to improve the classification accuracy of point cloud data, specifically focusing on the Actueel Hoogtebestand Nederland (AHN) dataset. The framework integrates geospatial knowledge into the deep learning process, enabling the model not only to refine its predictions through iterative learning but also to enhance the training data along the way via iterative online learning, ensuring continuous improvement in both training data quality and model performance.
The preprocessing phase assigns confidence scores to each point in the point cloud based on local neighborhood properties, with additional input from multispectral imagery (MSI) to further enhance the confidence estimation. These confidence scores are central to the online learning process, where the model prioritizes high-confidence points for training while progressively updating lower-confidence points to improve accuracy. To test the hypothesis that confidence-aware learning can enhance point cloud classification, we selected the KPConv network due to its suitability for handling unstructured data and capturing complex geometric features.
Extensive experiments demonstrate that the proposed framework, particularly with the Online strategy, enables deep learning models to perform better when trained solely on native point cloud attributes (elevation and intensity) compared to models without this strategy. Importantly, the Online strategy qualitatively enhances the training data by refining labels and reducing noise, thereby supporting more robust model performance. While incorporating additional features from aerial imagery showed no overall improvement, specific classes, like High tension and others did see performance gains. ...
The preprocessing phase assigns confidence scores to each point in the point cloud based on local neighborhood properties, with additional input from multispectral imagery (MSI) to further enhance the confidence estimation. These confidence scores are central to the online learning process, where the model prioritizes high-confidence points for training while progressively updating lower-confidence points to improve accuracy. To test the hypothesis that confidence-aware learning can enhance point cloud classification, we selected the KPConv network due to its suitability for handling unstructured data and capturing complex geometric features.
Extensive experiments demonstrate that the proposed framework, particularly with the Online strategy, enables deep learning models to perform better when trained solely on native point cloud attributes (elevation and intensity) compared to models without this strategy. Importantly, the Online strategy qualitatively enhances the training data by refining labels and reducing noise, thereby supporting more robust model performance. While incorporating additional features from aerial imagery showed no overall improvement, specific classes, like High tension and others did see performance gains. ...
Accurately classifying laser-scanned point cloud data remains a critical challenge in geospatial analysis, particularly due to the complexity and volume of the data. This thesis presents a novel, confidence-aware deep learning framework designed to improve the classification accuracy of point cloud data, specifically focusing on the Actueel Hoogtebestand Nederland (AHN) dataset. The framework integrates geospatial knowledge into the deep learning process, enabling the model not only to refine its predictions through iterative learning but also to enhance the training data along the way via iterative online learning, ensuring continuous improvement in both training data quality and model performance.
The preprocessing phase assigns confidence scores to each point in the point cloud based on local neighborhood properties, with additional input from multispectral imagery (MSI) to further enhance the confidence estimation. These confidence scores are central to the online learning process, where the model prioritizes high-confidence points for training while progressively updating lower-confidence points to improve accuracy. To test the hypothesis that confidence-aware learning can enhance point cloud classification, we selected the KPConv network due to its suitability for handling unstructured data and capturing complex geometric features.
Extensive experiments demonstrate that the proposed framework, particularly with the Online strategy, enables deep learning models to perform better when trained solely on native point cloud attributes (elevation and intensity) compared to models without this strategy. Importantly, the Online strategy qualitatively enhances the training data by refining labels and reducing noise, thereby supporting more robust model performance. While incorporating additional features from aerial imagery showed no overall improvement, specific classes, like High tension and others did see performance gains.
The preprocessing phase assigns confidence scores to each point in the point cloud based on local neighborhood properties, with additional input from multispectral imagery (MSI) to further enhance the confidence estimation. These confidence scores are central to the online learning process, where the model prioritizes high-confidence points for training while progressively updating lower-confidence points to improve accuracy. To test the hypothesis that confidence-aware learning can enhance point cloud classification, we selected the KPConv network due to its suitability for handling unstructured data and capturing complex geometric features.
Extensive experiments demonstrate that the proposed framework, particularly with the Online strategy, enables deep learning models to perform better when trained solely on native point cloud attributes (elevation and intensity) compared to models without this strategy. Importantly, the Online strategy qualitatively enhances the training data by refining labels and reducing noise, thereby supporting more robust model performance. While incorporating additional features from aerial imagery showed no overall improvement, specific classes, like High tension and others did see performance gains.
Pointcloud based anatomy
Synthesis project report
Student report
(2023)
-
G.D. Brouwer, S.C. Madanu, S.A. Epema, V. Bhatt, Q. Shen, P.J.M. van Oosterom, E. Verbree
The declining availability of practical hours for medical anatomical education has prompted Enatom to develop a digital anatomical platform, utilizing the open-source WebGL-based point cloud renderer Potree. This platform, which employs detailed point cloud scans of anatomical structures, aims to offer a dynamic and interactive educational experience. Although Enatom's focus is not directly on geomatics, the techniques employed in this project have strong parallels with those used in geomatics, thereby enabling a symbiotic exchange of expertise. This interdisciplinary approach enhances the development of Enatom’s digital platform, with the potential to contribute to the field of geomatics. To address existing user experience challenges, this project has added a lasso-selection tool tailored for Potree, advanced annotation capabilities, and methods for sensitive data anonymization within the point cloud. The project's outcomes will be available in an open-source format at https://github.com/GEO1101-Synthesis-Group4/Selection-Annotation-Repo. This project exemplifies the versatile application of geomatics expertise beyond its traditional scope, demonstrating its potential in enhancing diverse domains such as medical education.
...
The declining availability of practical hours for medical anatomical education has prompted Enatom to develop a digital anatomical platform, utilizing the open-source WebGL-based point cloud renderer Potree. This platform, which employs detailed point cloud scans of anatomical structures, aims to offer a dynamic and interactive educational experience. Although Enatom's focus is not directly on geomatics, the techniques employed in this project have strong parallels with those used in geomatics, thereby enabling a symbiotic exchange of expertise. This interdisciplinary approach enhances the development of Enatom’s digital platform, with the potential to contribute to the field of geomatics. To address existing user experience challenges, this project has added a lasso-selection tool tailored for Potree, advanced annotation capabilities, and methods for sensitive data anonymization within the point cloud. The project's outcomes will be available in an open-source format at https://github.com/GEO1101-Synthesis-Group4/Selection-Annotation-Repo. This project exemplifies the versatile application of geomatics expertise beyond its traditional scope, demonstrating its potential in enhancing diverse domains such as medical education.