Critical to computer vision applications, deep learning demands a massive volume of training data for great performance. However, encrypting the sensitive information in a photograph is fraught with difficulty, despite rapid technological advancements. The Advanced Encryption System (AES) is the bedrock of classical encryption technologies. The Data Encryption Standard (DES) has low sensitivity, with weak anti-hacking capabilities. In a chaotic encryption system, a chaotic logistic map is employed to generate a key double logistic sequence, and deoxyribonucleic acid (DNA) matrices are created by DNA coding. The XOR operation is carried out between the DNA sequence matrix and the key matrix. Finally, the DNA matrix is decoded to obtain an encrypted image. Given that encrypted images are susceptible to attacks, a rapid and efficient Convolutional Neural Network (CNN) denoiser is used that enhances the robustness of the algorithm by maximizing the resolution of rebuilt images. The use of a key mixing percentage factor gives the proposed system vast key space and great key sensitivity. Its implementation is examined using statistical techniques such as histogram analysis, information entropy, key space analysis and key sensitivity. Experiments have shown that the suggested system is secure and robust to statistical and noise attacks.

In this paper the image processing method is used to enhance the pneumonia bacteria images. This paper recognized the bacteria images based on two domains. The enhancement techniques used for bacteria image enhancement were median filter, wiener filter, single scale retinex and multiscale retinex. Image enhancement has a very important role in digital image processing. The median and wiener filters were used for grayscale image enhancement. Then single scale retinex and multiscale retinex were used for color image enhancement. Based on performance metrics identified median filter is suitable for bacteria images in grayscale image enhancement and multiscale retinex is suitable for bacteria color image enhancement (Tab. 2, Fig. 8, Ref. 21).

Action Recognition plays a vital role in many secure applications. The objective of this paper is to identify actions more accurately. This paper focuses on the two stream network in which keyframe extraction method is utilized before extracting spatial features. The temporal features are extracted using Attentive Correlated Temporal Feature (ACTF) which uses Long Short Term Memory (LSTM) for deep features. The spatial and temporal features are fused and classified using multi Support Vector Machine (multiSVM) classifier. Experiments are done on HMDB51 and UCF101 datasets. The results of the proposed method are compared with recent methods in terms of accuracy. The proposed method is proved to work better than other methods by achieving an accuracy of 96% for HMDB51 dataset and 98% for UCF101 dataset.

No abstract

No abstract