Survey on deep learning applications in digital image security

The image is a function of 2 (two) dimensions f (x, y), where x and y are the coordinates that show the space, and the value of f (x, y) is called the image intensity at the coordinates. Digital image is an image that has a limited value of x, y, f (x, y). Digital images consist of a limited number of elements, where each element has coordinates and values. These elements are called pixels (Gonzalez and Woods., 2007). One method of randomization in the image is Arnold Transformation where the image is transformed using a cat map, which is a 2x2 matrix. After transformation, the value of each pixel gets new coordinates until the image does not have its original shape. As a result of the above problems, the author uses the basis of Arnold Transformation, to change the value of the map paint with Fibonacci and Lucas series for designing this application so that the randomization results have a higher level of security. The purpose of this study is (1) to do digital image randomization using Fibonacci and Lucas transforms so that the image does not have its original shape anymore and (2) save the randomized images into storage memory. The benefits of this study are (1) the design of digital image randomization applications, so that image security increases and (2) the effectiveness of the Fibonacci and lucas transformation methods in digital image randomization.Keywords: Image Processing, Fibonacci and Lucas

Download Full-text

Research on Steganalysis of Digital Image Based on Deep Learning

10.1109/aemcse51986.2021.00114 ◽

2021 ◽

Author(s):

Pei Li ◽

Yeli Li ◽

Hongjuan Wang ◽

Chang Liu

Keyword(s):

Deep Learning ◽

Digital Image

Download Full-text

Research on Water Seepage Detection Technology of Tunnel Asphalt Pavement Based on Deep Learning and Digital Image Processing

10.21203/rs.3.rs-1235518/v1 ◽

2022 ◽

Author(s):

Jiaqi Li ◽

Zhaoyi He ◽

Dongxue Li ◽

Aichen Zheng

Keyword(s):

Deep Learning ◽

Network Model ◽

Digital Image ◽

Traffic Safety ◽

Asphalt Pavement ◽

Threshold Segmentation ◽

Water Seepage ◽

Detection Technology ◽

Seepage Characteristics ◽

The Impact

Abstract In order to improve the traffic safety of the tunnel pavement and reduce the impact of water seepage on the pavement structure, a convolutional neural network (CNN) model is established based on image detection technology to realize the identification, classification and statistics of pavement seepage. First, compared with the MobileNet network model, the deep learning model EfficientNet network model was built, and the accuracy of the two models was analyzed for pavement seepage recognition. The F1 Score was introduced to evaluate the accuracy and comprehensive performance of the two models for different types of seepage characteristics. Then the three gray processing methods, six threshold segmentation methods, as well as three filtering methods were compared to extract water seepage characteristics of digital image. Finally, based on the processed image, a calculation method of water seepage area was proposed to identify the actual asphalt pavement water seepage. The result shows that the recognition accuracy of the EfficientNet network model in the training set and the validation set are 99.85% and 97.53%, respectively, and the prediction accuracy is 98.00%. The accuracy of pavement water seepage recognition and prediction is better than the MobileNet network model. Using the cvtColor function for gray processing, using THRESH_BINARY for threshold segmentation, and using a combination of median filtering and morphological opening operations for image noise reduction can effectively extract water seepage characteristics. The water seepage area calculated by the proposed method has a small difference with the actual water seepage area, and the effect is agreeable.

Download Full-text