- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
[IEEE 2018 International Conference on 3D Vision (3DV) - Verona (2018.9.5-2018.9.8)] 2018 International Conference on 3D Vision (3DV) - Semantic Classification of 3D Point Clouds with Multiscale Spherical Neighborhoods
摘要： This paper introduces a new de?nition of multiscale neighborhoods in 3D point clouds. This de?nition, based on spherical neighborhoods and proportional subsampling, allows the computation of features with a consistent geometrical meaning, which is not the case when using k-nearest neighbors. With an appropriate learning strategy, the proposed features can be used in a random forest to classify 3D points. In this semantic classi?cation task, we show that our multiscale features outperform state-of-the-art features using the same experimental conditions. Furthermore, their classi?cation power competes with more elaborate classi?cation approaches including Deep Learning methods.
关键词： random forest,3D point clouds,multiscale spherical neighborhoods,deep learning,semantic classification
更新于2025-09-04 15:30:14
-
Variability of Multispectral Lidar 3D and Intensity Features with Individual Tree Height and Its Influence on Needleleaf Tree Species Identification
摘要： Tree species identification is important in forest management. The multispectral lidar Titan of Teledyne Optech Inc. can improve tree species separation by providing classification features computed from the three-channel intensities, ratios and normalized differences. However, the value of features used in classification algorithms (e.g., random forest, RF) may vary with tree size. The focus of the present study is to show how tree height influences the 3D and intensity features, how this relationship may affect the species classification accuracy, and how different classification strategies may circumvent this problem. Six needleleaf species (Pinus resinosa, Pinus strobus, Pinus sylvestris, Larix laricina, Picea abies and Picea glauca), found in plantations of different ages, were sampled to train classifiers. Some features yielded a good discriminatory power for species identification, despite their relation to tree height (r2 up to 0.6). Two classification strategies—a) using only size-invariant features (SIF) and b) training separate classifiers per tree height strata (HSC)—were compared to a standard classification (STD: all features, without height stratification). The accuracy of the SIF approach was lowest, useful variables being removed due to their relationship to tree height. The HSC provided only a minor improvement over the STD results.
关键词： tree species identification,Teledyne Optech Inc.,Titan,random forest,3D features,intensity features,multispectral lidar
更新于2025-09-04 15:30:14
-
[IEEE 2018 IEEE 7th Global Conference on Consumer Electronics (GCCE) - Nara, Japan (2018.10.9-2018.10.12)] 2018 IEEE 7th Global Conference on Consumer Electronics (GCCE) - Histogram-Based Image Pre-processing for Machine Learning
摘要： This paper proposes to use some image processing methods as a data normalization method for machine learning. Conventionally, z-score normalization is widely used for pre-processing of data. In the proposed approach, in addition to z-score normalization, a number of histogram-based image processing methods such as histogram equalization are applied to training data and test data as a pre-processing method for machine learning. We evaluate the effectiveness of the proposed approach by using a support vector machine algorithm and a random forest one. In experiments, the proposed scheme is applied to a face-based authentication algorithm with SVM/random forest classifiers to confirm the effectiveness. For SVM classifiers, both z-score normalization and image enhancement work well as a pre-processing method for improving the accuracy. In contrast, for random forest classifiers, a number of image enhancement methods work well, although z-score normalization is unuseful for improving the accuracy.
关键词： Support Vector Machines,Pre-processing,Contrast Enhancement,Random Forest,Machine Learning
更新于2025-09-04 15:30:14
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Automatic Mapping of Irrigated Areas in Mediteranean Context Using Landsat 8 Time Series Images and Random Forest Algorithm
摘要： Groundwater withdrawals by farmers, in Morocco, are very numerous and informal. Therefore, the need for information on the location of irrigated areas is becoming increasingly important. Our main objective, in this study, is to evaluate the use of high-resolution Landsat 8 (L8) time series images and Random forest (RF) method to produce a land cover map with a sufficient precision to monitor the extension of irrigated areas. In the first part of this study, four parameters were evaluated: Number of trees, min split samples, max features and max depth. The results proves that the last parameter is the most important and has more impact on the oob score, which can reach 91%. The second part of this study was devoted to reduce furthermore the number of features taken as input in the classification process. This was done through feature reduction then selection. The computational time was highly reduced and the best level of classification accuracy was reached by using only Landsat 8 (L8) time series images, statistics on the temporal spectral indices (NDVI, MNDWI) and Range texture.
关键词： Random forest,NDVI,Landsat 8,Irrigated areas,Feature selection,time series,MNDWI,tuning,Range texture
更新于2025-09-04 15:30:14
-
[IEEE 2018 IEEE International Conference on Intelligent Transportation Systems (ITSC) - Maui, HI, USA (2018.11.4-2018.11.7)] 2018 21st International Conference on Intelligent Transportation Systems (ITSC) - Machine Learning-based Stereo Vision Algorithm for Surround View Fisheye Cameras
摘要： Recently, automated emergency brake systems for pedestrian have been commercialized. However, they cannot detect crossing pedestrians when turning at intersections because the field of view is not wide enough. Thus, we propose to utilize a surround view camera system becoming popular by making it into stereo vision which is robust for the pedestrian recognition. However, conventional stereo camera technologies cannot be applied due to fisheye cameras and uncalibrated camera poses. Thus we have created the new method to absorb difference of the pedestrian appearance between cameras by machine learning for the stereo vision. The method of stereo matching between image patches in each camera image was designed by combining D-Brief and NCC with SVM. Good generalization performance was achieved by it compared with individual conventional algorithms. Furthermore, feature amounts of the point cloud reconstructed by the stereo pairs are utilized with Random Forest to discriminate pedestrians. The algorithm was evaluated for the actual camera images of crossing pedestrians at various intersections, and 96.0% of pedestrian tracking rate with high position detection accuracy was achieved. They were compared with Faster R-CNN as the best pattern recognition technique, and our proposed method indicated better detection performance.
关键词： NCC,automated emergency brake systems,machine learning,SVM,Faster R-CNN,stereo vision,pedestrian detection,D-Brief,Random Forest,surround view camera system
更新于2025-09-04 15:30:14
-
Hard exudate based severity assessment of diabetic macular edema from retinal fundus images
摘要： Diabetic macular edema (DME) is a consequence of diabetic retinopathy characterised by the abnormal accumulation of fluid and protein deposit in the macula region of the retina. Prior disclosure of even a trivial trace of DME is essential as it could consequently lead to blurred vision. DME can be diagnosed by the presence of exudates (glossy lesions) in the retinal fundus images. In this work, OD and macula are detected using morphological operation and hard exudates are segmented. Exudates are classified using early treatment diabetic retinopathy standard as normal, moderate or severe cases. The proposed work also incorporates the extraction of various features from the retinal fundus image. Various textural and exudate features are extracted and fed to a classifier to detect DME. Experiments are performed on a publically available database. Performance is evaluated with metrics like accuracy, sensitivity, specificity and accuracy. The results obtained are promising.
关键词： DME,random forest,hard exudates,diabetic macular edema,classification,macula,feature extraction,optic disc
更新于2025-09-04 15:30:14