Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A contour detection and heuristic technology, applied in the field of image processing, can solve problems such as inability to achieve universality, lack of contour information, and failure to ensure the integrity of the target contour.
Active Publication Date: 2021-11-19
GUANGXI UNIVERSITY OF TECHNOLOGY
View PDF9 Cites 0 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0004] Biological method: simply simulate part of the physiological characteristics of the visual system. In the process of contour extraction, some dynamic characteristics of vision cannot be simulated well, resulting in the lack of contour information and the enhancement of texture information to a certain extent. This kind of problem can not guarantee the integrity of the target contour
[0005] The convolutional neural network applied to computational vision tasks is not well integrated with the visual mechanism, and most of the traditional bionic algorithms use formulas to simulate a certain function of cells, which cannot achieve universality
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 more
Image
Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0041] A bioinspired multi-level multi-level feedback contour detection method provided in this embodiment includes the following steps:
[0042] A. Construct a deep neural network structure. The specific structure of the deep neural network is as follows:
[0043] Encoding network, decoding network;
[0044] Among them, the encoding network includes VGG16, a preprocessing module P, and a feedback network;
[0045] The VGG16 network is obtained by discarding the 3 fully connected layers and the last downsampling layer from the original VGG16 network, and is divided into five stages with the pooling layer as the dividing line; the preprocessing module P corresponds to the five stages in the VGG16 network There are 5;
[0046] The feedback network is provided with four horizontal feedback stages of L1, L2, L3, and L4 connected in sequence, four feedback modules F are set in the L1 horizontal feedback stage, three feedback modules F are set in the L2 horizontal feedback stage, ...
Embodiment 2
[0066] For the quantitative performance evaluation of the final contour map, we use the same performance measurement standard as in literature 1, and the specific evaluation is shown in formula (1).
[0067]
[0068] Among them, P represents the precision rate, and R represents the recall rate. The larger the value of F, the better the performance.
[0069] Document 1: Deng R, Liu S. Deep Structural Contour Detection[C] / / Proceedings of the 28th ACM International Conference on Multimedia.2020:304-312.
[0070] The parameters used in Document 1 are the same as the original text, and they are guaranteed to be the optimal parameters of the model.
[0071] Figure 5 The four natural images randomly selected from the Berkeley Segmentation Dataset (BSDS500), the corresponding real contour maps, the optimal contour map detected by the method in Document 1, and the optimal contour detected by the method in this paper are shown from left to right.
[0072] The performance comparis...
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 aims to provide a biological heuristic multi-stage multi-level feedback contour detection method. The method comprises the following steps of: constructing a deep neural network structure which comprises a coding network and a decoding network, wherein the coding network comprises a VGG16, a preprocessing module P and a feedback network; the VGG16 network takes a pooling layer as a boundary and is divided into five stages; preprocessing modules P are arranged corresponding to five stages in the VGG16 network; the feedback network is provided with four transverse feedback stages which are connected in sequence; the decoding network comprises a plurality of feedback modules F and an addition layer; and an original image passes through the coding network and the decoding network in sequence, and a final output contour is obtained. The completeness of the target contour is ensured, and the problem of discontinuous contour can be effectively solved.
Description
technical field [0001] The invention relates to the field of image processing, in particular to a bioinspired multi-level and multi-level feedback contour detection method. Background technique [0002] Contour detection aims to extract the dividing line between the background and the target in an image. It is usually used as a key step in the front-end processing of various intermediate and advanced computer vision tasks, and is one of the basic tasks in the field of computer vision research. Currently, contour detection has the following two methods: [0003] Deep learning method: Usually, public VGG-Net, Res-Net and other models are used for transfer learning as the encoding network representation features, and then research and design a matching decoding network to analyze the features and finally obtain the target contour. The convolutional neural network itself is inspired by biological mechanisms, but it has not been well integrated with it in later developments. ...
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
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More