A Fracture Image Compression Sampling Method Based on Generative Adversarial Network

A technology of compressed sampling and image compression, applied in biological neural network models, image communication, neural learning methods, etc., can solve the problems of reconstruction accuracy, noise robustness, reconstruction speed superiority, etc., to reduce discomfort Qualitative, data compression rate improvement, and the effect of saving energy consumption

CN111711820BActive Publication Date: 2021-03-30HARBIN INST OF TECH
5 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Publication Date
2021-03-30

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The present invention proposes a crack image compression sampling method based on a generative adversarial network. The method includes the network architecture design of the generative adversarial network, the crack image generator modeling of the mapping relationship between the crack image and the low-dimensional vector, and the hyperparameters of the adversarial training. Tuning, design of compressed observation matrix for compressed sampling, solution of optimal low-dimensional vector, etc. The method of the present invention adopts the crack image generator trained to generate an adversarial network as a physical constraint to realize the decompression and reconstruction of the image, without requiring the sparsity of the crack image as in the traditional compression sampling method, and has a wider application range. After the generative confrontation network learns the mapping relationship between the crack image and the low-dimensional vector, the low-dimensional vector is optimized based on the gradient descent method, and the fast solution of image decompression and reconstruction is realized. The method has unique advantages in the reconstruction accuracy and reconstruction speed of the fracture image under a relatively high compression rate, and has strong robustness to noise.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention belongs to the technical field of signal processing and structural health monitoring, in particular to a crack image compression sampling method based on a generative confrontation network. Background technique

[0002] At present, under the influence of long-term loads, environmental erosion and other factors, various types of infrastructure will inevitably be damaged. The continuous accumulation and development of damage will lead to the continuous decline of the bearing capacity and use function of the structure, until the safe use of the structure is endangered. Therefore, real-time monitoring of structural damage and evaluation of structural health through theoretical analysis are one of the core issues in structural health monitoring. The cracks on the surface of the structure are frequently monitored indicators, which can reflect the degree of damage to the structure and have a serious impact on the function of the structure. For ...

Examples

Embodiment

[0052] to combine image 3 , data compression is performed on crack images in three different backgrounds, and the image compression sampling method based on generative confrontation network of the present invention is used to decompress and reconstruct crack images.

[0053] The crack image resolution used is 128 pixels×128 pixels, the data is compressed by 16 times, and the measurement noise level of 5% is considered in the compressed data.

[0054] The crack image compression sampling method based on the generated confrontation network in the present invention is used to decompress and reconstruct:

[0055] The first step is as follows: collect a certain amount of high-resolution crack images on structural surfaces with different backgrounds, cut out the blocks with cracks in the images and uniformly scale the resolution to 128 pixels × 128 pixels, and create various cracks Large dataset of images.

[0056] The second step is specifically: after obtaining the above data s...