In today's digital age, securing sensitive visual data is crucial, particularly in fields like medical imaging and secure communication. This paper presents a novel greyscale image encryption algorithm that combines a secure message-based key generation approach with a substitution-permutation network (SPN) for enhanced security. The algorithm uses a user-specified parameter 'N' an input image, and a secret message to derive cryptographic parameters via a SHA-256 string for a Henon map. The map generates two arrays, x and y matching the image dimensions, which are used to create an XOR-based mask. Before applying the mask, the image undergoes P-rounds of SPN transformations, involving substitution and permutation operations. These operations employ a logistic map for randomness, ensuring robust encryption. Decryption reverses these steps. The paper highlights the algorithm's computational efficiency and strong security features, making it suitable for diverse applications.

Abstract Images are a crucial component in contemporary data transmission. Numerous images are transmitted daily through the open-source network. This paper presents a multi-image encryption scheme that utilises flip-shift-rotate synchronous-permutation-diffusion (FSR-SPD) processes to ensure the security of multiple images in a single encryption operation. The proposed encryption technique distinguishes itself from current multi-image encryption methods by utilising SPD operation and rapid FSR-based pixel-shuffling and diffusion operation. The SPD is a cryptographic technique that involves the simultaneous application of permutation and diffusion methods. The FSR-based process involves the manipulation of pixels through three different operations, namely flipping, shifting, and rotating. In the process of encryption, the image components of red, green, and blue colours are merged into a single composite image. The large image is partitioned into non-overlapping blocks of uniform size. The SPD technique is employed to tackle each specific block. The encryption method is efficient and expeditious as it exhibits high performance with both FSR and SPD procedures. The method employs a single, fixed-type, one-dimensional, piecewise linear chaotic map (PWLCM) for both the permutation and diffusion phases, resulting in high efficiency in both software and hardware. The proposed method is assessed using key space, histogram variance, neighbouring pixel correlation, information entropy, and computational complexity. The proposed method has a much bigger key space than the comparative method. Compared to comparison approaches, the suggested solution reduces encrypted picture histogram variance by 6.22% and neighbouring pixel correlations by 77.78%. Compared to the comparison technique, the proposed scheme has a slightly higher information entropy of 0.0025%. Other multiple-color image encryption methods are more computationally intensive than the suggested method. Computer simulations, security analysis, and comparison analysis evaluated the proposed methodology. The results show it outperforms multiple images encrypting methods.

This paper proposes an image securing technique that aims to provide two-level security on two images in terms of encryption at the same time. In this technique, both bit and pixel-level encryptions are carried out; first, the pixel-level-shuffling is performed using the piece-wise linear chaotic map (PWLCM); then, the diffusion in bit-level is performed using the key-image. The bit-level diffusion using chaos not just to confuse the pixels, but is also diffuses them intensely. In addition, the bit and pixel-level processes improve that algorithm's security. Additionally, the parallel bit-plane diffusion process reduces the method's computational complexity. This technique uses one type of one-dimensional chaotic map in both permutation and diffusion, thereby increasing the algorithm's hardware and software efficiency. The results of the security analysis and simulation show that the suggested method is more effective in encoding and improves the security of the encrypted images.

Body pose analysis is an important factor of human action recognition. Recently, the proposed Recurrent Neural Networks (RNNs) and deep ConvNets-based methods are showing good performances in learning sequential information. Despite these good performances, RNN lacks to efficiently learn spatial relation between body parts while deep ConvNets require a huge amount of data for training. We propose a Distance-based Neural Network (DNN) for action recognition in static images. We compute effective distances between a set of body part pairs for a given image and feed to DNN to learn effective representation of complex actions. We also propose Distance-based Convolutional Neural Network (DCNN) to learn representations from 2D images. The distances are rearranged in 2D grayscale image called as a Distance Image. This 2D representation allows the network to learn specific discriminative information between adjacent pixel distance values corresponding to different body part pairs. We evaluate our method on two real-world datasets i.e. UT-Interaction and SBU Kinect Interaction. Results show that our proposed method achieves better performance compared to the state-of-the-art approaches.

Abstract Most interaction recognition approaches have been limited to single‐person action classification in videos. However, for still images where motion information is not available, the task becomes more complex. Aiming to this point, we propose an approach for multiperson human interaction recognition in images with keypoint‐based feature image analysis. Proposed method is a three‐stage framework. In the first stage, we propose feature‐based neural network (FCNN) for action recognition trained with feature images. Feature images are body features, that is, effective distances between a set of body part pairs and angular relation between body part triplets, rearranged in 2D gray‐scale image to learn effective representation of complex actions. In the later stage, we propose a voting‐based method for direction encoding to anticipate probable motion in steady images. Finally, our multiperson interaction recognition algorithm identifies which human pairs are interacting with each other using an interaction parameter. We evaluate our approach on two real‐world data sets, that is, UT‐interaction and SBU kinect interaction. The empirical experiments show that results are better than the state‐of‐the‐art methods with recognition accuracy of 95.83% on UT‐I set 1, 92.5% on UT‐I set 2, and 94.28% on SBU clean data set.