Deep capsule network image classification method and system based on adaptive spatial mode
A technology of spatial patterns and network images, applied in neural learning methods, character and pattern recognition, biological neural network models, etc., can solve the problems of inaccurate classification accuracy of hyperspectral image classification datasets and complex spatial structure characteristics.
Pending Publication Date: 2021-05-07
SUN YAT SEN UNIV +1
View PDF8 Cites 2 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] In order to solve the above-mentioned technical problems, the object of the present invention is to provide a deep capsule network image classification method and system based on adaptive spatial patterns to solve the problem of inaccurate classification accuracy of hyperspectral image classification data sets with complex spatial structure features
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
[0044] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. For the step numbers in the following embodiments, it is only set for the convenience of illustration and description, and the order between the steps is not limited in any way. The execution order of each step in the embodiments can be adapted according to the understanding of those skilled in the art sexual adjustment.
[0045] The present invention constructs a group of self-adaptive convolution units (ASP units), and on the basis of expanding the receptive field, the shape adaptively adjusts the learning area of the convolution units, and provides a solution for the classification of hyperspectral image data sets with complex texture structures . That is, only by adding a small amount of transformation parameters inside the convolution unit, more excellent image detail features can be learned. Specifically, the present invention...
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 discloses a deep capsule network image classification method and system based on an adaptive spatial mode, and the method comprises the steps of carrying out the image preprocessing of a training image, and constructing a training set; based on the training set, carrying out adaptive space unit deep capsule network training to obtain a trained deep capsule network; obtaining a to-be-detected image and inputting the to-be-detected image into the trained deep capsule network for image classification to obtain an image classification result, wherein the deep capsule network comprises an ASPConvs module, an ASPCaps module and a full-connection capsule layer. The system comprises a preprocessing module, a training module and a classification module. According to the invention, the method achieves the adaptive learning of complex object structure information, and improves the image classification precision of a dense texture region. The deep capsule network image classification method and system based on the adaptive spatial mode can be widely applied to the field of image classification.
Description
technical field [0001] The present invention relates to the field of image classification, in particular to a deep capsule network image classification method and system based on an adaptive spatial pattern. Background technique [0002] Traditional deep learning image classification algorithms include: stacked autoencoder network, deep confidence neural network, recurrent neural network, generation of confrontational neural network, and convolutional neural network, among which traditional neural network-based image classification methods pass convolution, pooling Image feature extraction is carried out through steps such as normalization and fully connected layers. However, this type of method often requires a large number of data samples for network model training, and the training process is complex, and it is difficult to extract key points of interest in images under complex scene features. And for the traditional neural network method, the position of the sampling po...
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): G06K9/62G06N3/08
CPCG06N3/084G06F18/213G06F18/2415G06F18/214Y02A40/10
Inventor 王锦萍李军谭晓军黄力陈霞
Owner SUN YAT SEN UNIV