56,801 research outputs found
Automatic Multi-view Surface Matching
No full textIn this paper we tackle the problem of automatically aligning an unordered set of range views. We propose a full pipeline that goes from the scans to the complete 3D model. The emphasis is on the automation – no manual in- tervention is require – and on the fact that no knowledge on the acquisition sequence is assumed. The contribution is twofold: in the pre-alignment phase a voting scheme is proposed that discovers the overlapping relationship among views; in the final refinement step we extend the Levenberg Marquardt-ICP to work with multiple views, in order to solve for the absolute pose of all images simultaneously
Visual vocabulary signature for 3D object retrieval and partial matching
No full textIn this paper a novel object signature is proposed for 3D object retrieval and partial matching. A part-based representation is obtained by partitioning the objects into subparts and by characterizing each segment with different geometric descriptors. Therefore, a Bag ofWords framework is introduced by clustering properly such descriptors in order to define the so called 3D visual vocabulary. In this fashion, the object signature is defined as a histogram of 3D visual word occurrences. Several examples on the Aim@Shape watertight dataset demonstrate the versatility of the proposed method in matching either 3D objects with articulated shape changes or partially occluded or compound objects. In particular, a comparison with the methods that participated to the Shape Retrieval contest 2007 (SHREC) reports satisfactory results for both object retrieval and partial matching
Registration of multiple acoustic range views for underwater scene reconstruction
No full textThis paper proposes a technique for the three-dimensional reconstruction of an underwater environment from multiple acoustic range views acquired by a remotely operated vehicle. The problem is made challenging by the very noisy nature of the data, the low resolution and the narrow field of view. Our main contribution is a new global registration technique to distribute registration errors evenly across all views. Our approach does not use data points after the first pairwise registration, for it works only on the transformations. Therefore, it is fast and occupies only a small memory. Experimental results suggest the global registration technique is effective in equalizing the error. Moreover, we introduce a statistically sound thresholding (the X84 rejection rule) to improve ICP robustness against noise and non-overlapping data
Accurate and automatic alignment of range surfaces
No full textThis paper describes an automatic pipeline that is able to take a set of unordered range images and align them into a full 3D model. A global voting scheme is employed for view matching, inspired by 2D techniques for image mo- saicing. Then a multiple view registration approach is in- troduced, which aims at optimizing the alignment error si- multaneously for all the views. Experiments demonstrate the effectiveness of the method
The bag of words approach for retrieval and categorization of 3D objects
No full textIn this paper, we propose a novel framework for 3D object retrieval and categorization. The object is modeled in terms of its subparts as an histogram of 3D visual word occurrences. We introduce an effective method for hierarchical 3D object segmentation driven by the minima rule that combines spectral clustering-for the selection of seed-regions-with region growing based on fast marching. Descriptors attached to the regions allow the definition of the visual words. After coding of each object according to the Bag-of-Words paradigm, retrieval can be performed by matching with a suitable kernel, or categorization by learning a Support Vector Machine. Several examples on the AimShape watertight dataset and on the Tosca dataset demonstrate the versatility of the proposed method in working with either 3D objects with articulated shape changes or partially occluded or compound objects. Results are encouraging as shown by the comparison with other methods for each of the analyzed scenarios
A bag of words approach for 3D object categorization
No full textIn this paper we propose a novel framework for 3D object categorization. The object is modeled it in terms of its sub-parts as an histogram of 3D visual word occurrences. We introduce an effective method for hierarchical 3D object segmentation driven by the minima rule that combines spectral clustering - for the selection of seed-regions - with region growing based on fast marching. The front propagation is driven by local geometry features, namely the Shape Index. Finally, after the coding of each object according to the Bag-of-Words paradigm, a Support Vector Machine is learnt to classify different objects categories. Several examples on two different datasets are shown which evidence the effectiveness of the proposed framewor
3D Acoustic Image Segmentation by a Ransac-Based Approach
No full textIn this paper, a new technique for 3D acoustic image segmentation and modelling is proposed. Especially, in the underwater environment, in which optical sensors suffer from visibility problems, the acoustical devices may provide efficient solutions, but, on the other hand, acoustic image interpretation is surely more difficult for a human operator. The proposed application involves the use of an acoustic camera which directly acquires images structured as a set of 3D points. Due to the noisy nature of this type of data, the segmentation problem becomes more challenging and the standard algorithms for range image segmentation are likely to fail. The proposed method is based on a simplified version of the so called recover and select paradigm in which the seed areas, from which the segmentation starts, are generated by adopting a robust approach based on the RANSAC (RANdom Sample And Consensus) algorithm. Superquadric primitives are directly recovered from raw data without any pre-segmentation processing. Experimental trials using real acoustical images confirm the goodness of the method, and a large robustness of the resulting segmented images, associated to a relatively low computational load
Supervised Learning of Diffusion Distance to Improve Histogram Matching
No full textn this paper we propose a learning method properly designed for histogram comparison. We based our approach on the so called diffusion distance which has been introduced to improve the robustness against the quantization effect and the limitations of the standard bin-to-bin distance computation. We revised the diffusion distance definition in order to cast the histogram matching as a distance metric learning problem. In particular, we exploit the Large Margin Nearest Neighbor (LMNN) classification procedure to introduce a supervised version of the standard nearest neighbor (NN) classification paradigm.
We evaluate our method on several application domains namely, brain classification, texture classification, and image classification. In all the experiments our approach shown promising results in comparison with other similar methods
View Synthesis From A Single Uncalibrated Image
No full textThis paper presents a method for generating synthetic views of a soccer ground starting from a single uncali- brated image. The relative affine structure of the players is computed by exploiting the knowledge of the soccer ground geometry and the fact that the players are in vertical positions. Then, novel views are generated using the “plane+parallax” representation to reproject points
- …
