Depth image clustering method based on Laplacian rank constraint
A deep image and clustering method technology, applied in the field of machine learning, can solve problems such as difficulty in obtaining good feature representations suitable for clustering, inability to handle large-scale data sets, and inability to handle a variety of different modal representation forms. The effect of solving high computational complexity problems, improving clustering accuracy, and reducing computational complexity
Pending Publication Date: 2022-02-11
NORTHWESTERN POLYTECHNICAL UNIV
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
The invention can obtain a good low-dimensional representation of data, and can flexibly process image data of different scales, and can solve the problem that the existing methods cannot handle large-scale data sets, it is difficult to obtain a good feature representation suitable for clustering, and it is difficult to process a variety of different Modal representations of data and other issues
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0019] The present invention will be further described below in conjunction with the accompanying drawings and embodiments, and the present invention includes but not limited to the following embodiments.
[0020] Such as figure 1 As shown, the present invention provides a depth image clustering method based on Laplacian rank constraints, and its specific implementation process is as follows:
[0021] 1. Dataset enhancement and preprocessing
[0022] Input the image data set, use rotation, scaling and color transformation to process the images in it, so as to expand the amount of data, enhance the expressive ability of the data set, and then perform size normalization processing to keep all images in the same size. Suppose the obtained image data set is X={x 1 ,x 2 ,...,x n}, where x i Indicates the i-th image in the data, i=1, 2,...,n, n indicates the total number of images contained in the expanded image data set, and there are k types of images in total.
[0023] 2. C...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention provides a depth image clustering method based on Laplacian rank constraint. The method comprises the following steps: firstly, preprocessing an image data set to obtain an expanded data set; secondly, introducing orthogonal constraint into a last layer of a spectrum embedded network to ensure that clustering indication vectors output by the network are mutually orthogonal, introducing a similarity matrix of Laplacian matrix rank limitation into a loss function, and training the network by using an image data set; finally, processing a to-be-classified image by using the trained network to obtain a classification result. According to the invention, good data low-dimensional representation can be obtained, the method is suitable for clustering processing of image data of different scales, a large-scale image data set can be efficiently processed, and the embodiment of the invention has good practical value.
Description
technical field [0001] The invention belongs to the field of machine learning, and in particular relates to a deep image clustering method based on Laplacian rank constraints. Background technique [0002] As a basic method in the field of machine learning, clustering has become more and more useful in the era of big data. Regardless of all walks of life, in the face of massive data, clustering is undoubtedly the lowest cost unsupervised data analysis method. Cluster analysis It has also become the primary tool for data analysis in many fields, including mathematics, computer science, statistics, biology, and economics. However, with the diversification of data forms, the existing clustering methods are somewhat stretched when dealing with multi-scale and complex manifold data. [0003] Traditional clustering methods include the spectral clustering method proposed by Ulrike von Luxburg in the literature "A Tutorial on SpectralClustering. Statistics and Computing, vol.17, no...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06V10/762G06V10/82G06K9/62G06N3/04G06N3/08
CPCG06N3/04G06N3/084G06F18/23213
Inventor 李学龙韦腾飞赵阳
Owner NORTHWESTERN POLYTECHNICAL UNIV