IEEE Matlab Projects in Chennai

Abstract :

Steganography is the process of hiding confidential messages in a digital image, video, or audio file, and avoiding distracting the traffic of secret information through the communication channel. Neural network-based steganography for information hiding (NIH) has existed presented. The cover image, also known as the carrier, is generally visible. In this paper, in-depth learning modules with LSB encryption, using the Adam algorithm, are used to teach the model, which includes a hidden network and an exposure network. The encoder neural network determines where and how to place the message, scattering it throughout the bits of the cover image. The decryption network on the receiving side is trained simultaneously with the encoder and reveals the secret image. The main feature of this work is that it creates minimal distortion to the secret message. This work has wide and secured applications in many domains. Finally, it has been shown that the proposed method has the best performance when using digital images with large changes in hue.

Abstract :

The texture edge continuity of a finger vein image is required for the accuracy of feature extraction. A finger vein image in the painting process with Gabor texture discretion is proposed. The proposed method virtually protects the texture edge continuity of the inpainted image. Firstly, utilizing the proposed vertical stage difference coding strategy, the Gabor texture feature matrix of the finger vein image, which can accurately represent the texture data, can be extracted from the Gabor filtering reactions. Then according to the local texture continuity of the finger vein image, the known pixels, which have different texture exposures with the center pixel in the patch, are filtered out using the Gabor texture constraining mechanism during the inpainting method. The suggested process destroys irrelevant information interference in the inpainting process and has a more precise texture propagation. Simulation investigations of artificially artificial images and acquired images show that the finger vein images inpainted by the proposed method have better texture continuity and higher image quality than the traditional methods. Which do not have Accurate texture constraints. The presented method enhances the recognition performance of the finger vein identification system with the accepted damaged images.

Abstract :

Advances in 5G technology, big data, and cloud storage have fueled the rapid growth of the Internet of Things (IoT). Based on the strict care requirements and the high level of accuracy required for diagnosis and pathological analysis, a big number of 3D medical module data has developed in robust watermarking. This feature integrates human visual features with enhanced cognitive hashing techniques. This is a powerful and efficient binary sequence. Zero embedding and blind extraction ensure that the original clinical volume data are not altered in any way, which meets the special requirements for diagnosis. With simulation results, the algorithm is robust and effectively resists common attacks and geometric attacks. Medical quantity data and watermark information used fewer Robust features to effectively bind the stored bandwidth and satisfy the security of the transfer and storage of clinical quantity data on the Internet of Things. In particular, the proposed algorithm increases the mean NC value under geometric attacks by 46.67% compared to sophisticated technology.

Abstract :

Face detection is critical for real-world applications such as video surveillance, human-machine communication, and security systems. This paper proposes a modified Convulsive Neural Network (CNN) structure by combining two normalization functions into two layers. The normalization function called the provided block normalization speeds up the network. With the continuous advancement of The Times and the advancement of technology, the rise of network social media has also brought about an “explosive” growth of image data. A series of algorithmic functions such as image eigenvalue extraction, authentication, and convolution is used to identify and analyze different images. Learn each data and use algorithms to predict the outcome. It has become a critical key to opening the door to artificial intelligence. From a mechanical point of view, image recognition is fundamental, but my problem with image recognition is how it relates the minimal amount of information in the image to high-level image semantics. The multi-level information fusion model based on the VGG16 model is an improvement over a fully integrated neural network. The convolutional neural network does not use the entire connection system for each layer of neurons in the neural network but uses certain nodes for connection. VGG divides the network into five groups (representing the five layers of Alexnet), but it uses 3 * 3 filters and connects them in a convolution sequence. The deeper the network DCNN, the larger the channel number.

Abstract :

Internet voting systems based on the E2E verification policy face many challenges; The most important is its safety. Over the past decade, many E2E voting systems have been discussed to analyze the e-voting system and formalize its security requirements. This article presents the E2E Verifiable Internet Voting System, which provides the voter with mobility and allows them to vote secretly on the public system for the benefit of early voting. The proposed system aims to globally support the electoral process using voter identification and biometric features. We propose a new identity-based blind signature scheme that ensures voter name anonymity. We accept the Boneh-Lynn-Shacham Short Signature Plan, which guarantees the privacy of the ballot with the lowest turnout. The Colombian government wants to implement electronic voting. However, the current electronic voting protocol has only a few security features required, and Colombia needs a protocol with all of these features to ensure a fraud-free election. We present the structure of the SIVP Secure Internet Voting Protocol, a new voting protocol for electoral procedures based on blind signatures and public-key cryptography.