Image retrieval is the process of retrieving images from a database. Certain algorithms have been used for traditional image retrieval. However, such retrieval involves certain limitations, such as manual image annotation, ineffective feature extraction, inability capability to handle complex queries, increased time required, and production of less accurate results. To overcome these issues, an effective image retrieval method is proposed in this study. This work intends to effectively retrieve images using a best feature extraction process. In the preprocessing of this study, a Gaussian filtering technique is used to remove the unwanted data present in the dataset. After preprocessing, feature extraction is applied to extract features, such as texture and color. Here, the texture feature is categorized as a gray level cooccurrence matrix, whereas the novel statistical and color features are considered image intensity-based color features. These features are clustered by k-means clustering for label formation. A modified genetic algorithm is used to optimize the features, and these features are classified using a novel SVMbased convolutional neural network (NSVMBCNN). Then, the performance is evaluated in terms of sensitivity, specificity, precision, recall, retrieval and recognition rate. The proposed feature extraction and modified genetic algorithm-based optimization technique outperforms existing techniques in experiments, with four different datasets used to test the proposed model. The performance of the proposed method is also better than those of the existing (RVM) regression vector machine, DSCOP, as well as the local directional order pattern (LDOP) and color co-occurrence feature + bit pattern feature (CCF + BPF) methods, in terms of the precision, recall, accuracy, sensitivity and specificity of the NSVMBCNN.

Data security can involve embedding hidden images, text, audio, or video files within other media to prevent hackers from stealing encrypted data. Existing mechanisms suffer from a high risk of security breaches or large computational costs, however. The method proposed in this work incorporates low-complexity encryption and steganography mechanisms to enhance security during transmission while lowering computational complexity. In message encryption, it is recommended that text file data slicing in binary representation, to achieve different lengths of string, be conducted before text file data masking based on the lightweight Lucas series and mod function to ensure the retrieval of text messages is impossible. The steganography algorithm starts by generating a random key stream using a hybrid of two low-complexity chaotic maps, the Tent map and the Ikeda map. By finding a position vector parallel to the input image vector, these keys are used based on the previously generated position vector to randomly select input image data and create four vectors that can be later used as input for the Lah transform. In this paper, we present an approach for hiding encrypted text files using LSB colour image steganography by applying a low-complexity XOR operation to the most significant bits in 24-bit colour cover images. It is necessary to perform inverse Lah transformation to recover the image pixels and ensure that invisible data cannot be retrieved in a particular sequence. Evaluation of the quality of the resulting stego-images and comparison with other ways of performing encryption and message concealment shows that the stego-image has a higher PSNR, a lower MSE, and an SSIM value close to one, illustrating the suitability of the proposed method. It is also considered lightweight in terms of having lower computational overhead.

No abstract

The continuous increase in the use of smart devices and the need for E2E smart2smart (S2S) services in IoT systems play effective and contemporary roles in the field of communication, and a large amount of resources is required. Thus, IoTs and cloud computing must be integrated. One of the results of this integration is the increase in the number of attacks and vulnerabilities in the E2E S2S message delivery service of such an IoT-cloud system. However, none of the traditional security solutions can be sufficiently regarded as a secure and lightweight mechanism for ensuring that the security requirements for E2E S2S message transmission in the IoT-cloud system are fulfilled. This work aims to provide an efficient and secure, lightweight E2E S2S message delivery function, which includes the E2E S2S secure key and biometric parameter exchange function, a bio-shared parameter and bio-key generation function, secure lightweight E2E S2S communication negotiation and secure E2E S2S lightweight message delivery. The secure, lightweight cryptographic communication procedure is negotiated between a pair of smart devices during each E2E session to minimize the power consumption required of limited-energy devices. Such a negotiation process prevents known attacks by providing responsive mutual authentication. Lightweight message delivery by the two smart devices can satisfy the basic security requirements of E2E communication and ensure that the computational cost required for a real-time system is as low as possible. INDEX TERMS Message delivery function, IoT-cloud system, smart devices, E2E S2S, mutual authentication.

Despite extensive research on content-based image retrieval, challenges such as low accuracy, incapability to handle complex queries and high time consumption persist. Initially, a preprocessing technique is introduced in this study, a technique that uses a median filter to remove noise to achieve improved accuracy and reliability. Then, Fourier and circularity descriptors are extract in an effective manner correspondent to the texture and affine shape adaptation features. In addition, various descriptors, such as color histogram, color moment, color autocorrelogram and color coherency vector, are extracted as the invariant color features. The multiple ant colony optimization (MACOBTC) approach is implemented with whole features to find relevant features. Finally, the relevant features are utilized for the greedy learning of deep Boltzmann machine classifier (GDBM). The proposed approach obtains effective performance and accurate results on four datasets and is analyzed with various parameters such as accuracy, precision, recall, Jaccard, Dice, and Kappa coefficients. The GDBM provides a 25% increase in accuracy compared with existing techniques, such as the a priori classification algorithm.

Background: the considerable time consumption, query retrieval difficulty and reduced retrieval rate. Still remaining challenges in Content-based image retrieval.Methods: in this work, we propose a pre-processing method that uses a Gaussian filter to improve quality by reducing image noise. An effective feature extraction method for in presented to extracted texture help color co-occurrence feature (CCF), color and shape features such as area and diameters. The colors features are extracted by means of a grey-level co-occurrence matrix and bit pattern. Extracting these features will enhance the image retrieval accuracy. With the use of a novel multi-SVM classifier, classification is performed and image retrieval is completed effectively.Results: performance measures, namely, precision, recall, error rate, correct rate, and retrieval rate, are computed. The proposed methodology produces superior results on these measures and exhibits an effective retrieval rate of approximately 94.92%; therefore, our technique is more efficient than existing MRED and MALP methods.Index Terms-Pre-processing, feature extraction, novel Multi-SVM classifier, color co-occurrence feature (CCF), grey level co-occurrence feature (GLCM), bit pattern feature (BPF).

No abstract