Abstract:<span>Since the skeleton represents the topology structure of the query sketch and 2D views of 3D model, this paper proposes a novel sketch-based 3D model retrieval algorithm which utilizes skeleton characteristics as the features to describe the object shape. Firstly, we propose advanced skeleton strength map (ASSM) algorithm to create the skeleton which computes the skeleton strength map by isotropic diffusion on the gradient vector field, selects critical points from the skeleton strength map and conn… Show more
“…In sketch-based algorithm, 3D model is projected into a set of 2D views. Zhang et al (2017) used PCA-DAISY descriptor and fisher coding algorithm to retrieve 3D model according to 2D sketch. Nie, Wang & Lu (2018) proposed Multi-Scale and Multi-Channel CNN to extract features for 3D model retrieval.…”
3D (three-dimensional) models are widely applied in our daily life, such as mechanical manufacture, games, biochemistry, art, virtual reality, and etc. With the exponential growth of 3D models on web and in model library, there is an increasing need to retrieve the desired model accurately according to freehand sketch. Researchers are focusing on applying machine learning technology to 3D model retrieval. In this article, we combine semantic feature, shape distribution features and gist feature to retrieve 3D model based on interactive attention convolutional neural networks (CNN). The purpose is to improve the accuracy of 3D model retrieval. Firstly, 2D (two-dimensional) views are extracted from 3D model at six different angles and converted into line drawings. Secondly, interactive attention module is embedded into CNN to extract semantic features, which adds data interaction between two CNN layers. Interactive attention CNN extracts effective features from 2D views. Gist algorithm and 2D shape distribution (SD) algorithm are used to extract global features. Thirdly, Euclidean distance is adopted to calculate the similarity of semantic feature, the similarity of gist feature and the similarity of shape distribution feature between sketch and 2D view. Then, the weighted sum of three similarities is used to compute the similarity between sketch and 2D view for retrieving 3D model. It solves the problem that low accuracy of 3D model retrieval is caused by the poor extraction of semantic features. Nearest neighbor (NN), first tier (FT), second tier (ST), F-measure (E(F)), and discounted cumulated gain (DCG) are used to evaluate the performance of 3D model retrieval. Experiments are conducted on ModelNet40 and results show that the proposed method is better than others. The proposed method is feasible in 3D model retrieval.
“…In sketch-based algorithm, 3D model is projected into a set of 2D views. Zhang et al (2017) used PCA-DAISY descriptor and fisher coding algorithm to retrieve 3D model according to 2D sketch. Nie, Wang & Lu (2018) proposed Multi-Scale and Multi-Channel CNN to extract features for 3D model retrieval.…”
3D (three-dimensional) models are widely applied in our daily life, such as mechanical manufacture, games, biochemistry, art, virtual reality, and etc. With the exponential growth of 3D models on web and in model library, there is an increasing need to retrieve the desired model accurately according to freehand sketch. Researchers are focusing on applying machine learning technology to 3D model retrieval. In this article, we combine semantic feature, shape distribution features and gist feature to retrieve 3D model based on interactive attention convolutional neural networks (CNN). The purpose is to improve the accuracy of 3D model retrieval. Firstly, 2D (two-dimensional) views are extracted from 3D model at six different angles and converted into line drawings. Secondly, interactive attention module is embedded into CNN to extract semantic features, which adds data interaction between two CNN layers. Interactive attention CNN extracts effective features from 2D views. Gist algorithm and 2D shape distribution (SD) algorithm are used to extract global features. Thirdly, Euclidean distance is adopted to calculate the similarity of semantic feature, the similarity of gist feature and the similarity of shape distribution feature between sketch and 2D view. Then, the weighted sum of three similarities is used to compute the similarity between sketch and 2D view for retrieving 3D model. It solves the problem that low accuracy of 3D model retrieval is caused by the poor extraction of semantic features. Nearest neighbor (NN), first tier (FT), second tier (ST), F-measure (E(F)), and discounted cumulated gain (DCG) are used to evaluate the performance of 3D model retrieval. Experiments are conducted on ModelNet40 and results show that the proposed method is better than others. The proposed method is feasible in 3D model retrieval.
“…Jing Zhang et al [7] have proposed a new sketch-based 3D model retrieval procedure which employs skeleton characteristics as the features to define the object shape.…”
A huge number of three-dimensional models exists on the internet, due to the fact that there are now more three-dimensional modelling and digitizing tools available for ever-increasing applications. The procedures for retrieval of three-dimensional models have thus become even more essential. The subject of this paper is a shape retrieval of 3D models that are signified as triangle meshes. We propose a new method which first computes the descriptor of 3D models through extracting its features, and then divides a model into clusters depending on a descriptor which is invariant to scale and orientation. A Fuzzy C-means clustering method is utilized for dividing the model into clusters. The superior performance and benefits of our method are shown in the results.
More accurate tree models, such as branch skeleton, are needed to acquire forest inventory data. Currently available algorithms for constructing a branch skeleton from a LiDAR point cloud have low accuracy with problems such as irrational connection near trunk bifurcation, excessive central deviation and topological errors. Using the C++ and PCL library, a novel algorithm of the incomplete simulation of tree transmitting water and nutrients (ISTTWN), based on geometric characteristics for tree branch skeleton extraction, was developed in this research. The algorithm is an incomplete simulation of tree transmitting water and nutrients. Improvements were made to improve the time and memory consumption. The result show that the ISTTWN algorithm without any improvements is quite time consuming but has consecutive output. After improvement with iteration, the process is faster and has more detailed output. Breakpoint connection is added to recover continuity. The ISTTWN algorithm with improvements can produce a more accurate skeleton and cost less time than a previous algorithm. The superiority and effectiveness of the method are demonstrated, which provides a reference for the subsequent study of tree modeling and a prospect of application in other fields, such as virtual reality, computer games and movie scenes.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.