Research On Digital Image Security Method Based On Advanced Encryption Technology

The internet,integral in today’s technology,facilitates various tasks,but its misuse poses inherent risks.Consequently,the protection of information has garnered heightened importance.However,challenges persist in securing and transferring,processing or storing information either online or offline.Recent studies on the cybersecurity of digital images highlight the efficacy of numerous security mechanisms addressing image-related concerns.The emergence of technology incorporating artificial intelligence introduces new vulnerabilities,enabling unauthorized access to information.Hackers exploit AI to decipher data transmitted,stored or processed through image formats.A notable challenge lies in the defense against deep learning cryptanalysis attacks.In image encryption,the emerging challenges stemming from encryption’s security,efficiency and quality.For encryption’s security,traditional approaches rely on a limited key space,exposing images to significant risks of information breaches through brute-force attacks,especially orchestrated by deep learning models.To enhance the encryption’s security,a proposed security mechanism is developed to effectively address this challenge by generating a larger number of key spaces.This approach aims to fortify image encryption,significantly enhancing resistance against unauthorized access and potential data breaches.Moreover,based on our observation,we find that the repercussions of computational complexity led to slow encryption and decryption processes,which significantly harm the encryption’s efficiency.To concur this challenge,developing a lightweight encryption mechanism is required.We propose our solution,which can feature an increased number of key spaces to speed up the overall encryption process.Finally,we observe that potential alterations or modifications of pixel values during the encryption process is an emerging vulnerability,especially during digit extraction or digit concatenation.This susceptibility may lead to a decrease in the quality of the current encryption.Accordingly,a mechanism is designed to effectively learn the characteristics of different types of images,which will improve the ultimate visual accuracy.By mitigating the risks associated with potential pixel value alterations,this approach ensures the preservation of visual quality and upholds data integrityThe proposed methods demonstrate remarkable effectiveness in image encryption.In the aspect of generating key space,a total number of 265536 space has been generated,surpassing the recommended standard of 2100.The execution time is notably efficient,clocking in at 0.0031 seconds with a key size of 65536 bits compared to 0.2148075 seconds with the average of 299 bits.Quality assessments indicate a flawless decryption process with negligible in terms of Mean Squared Error(MSE)and 0%Peak Signal-to-Noise Ratio(PSNR),showcasing resilience against various attacks.Based on the evaluations,by applying the techniques proposed,the encryption of the digital’s image’s security,efficiency and quality is significantly enhanced.The following are contribution of this dissertation.(1)For better encryption security,traditional methods often use a limited set of keys,making images vulnerable to information breaches.To address this,we’ve developed a new security mechanism.It enhances encryption by generating a larger number of unique keys using the Sigmoid Logistic Map and Kronecker XOR product.(2).The impact of computational complexity has resulted in slow encryption and decryption processes,significantly affecting the efficiency of encryption.To overcome this challenge,we propose a solution in the form of a lightweight encryption mechanism which use Fuzzy number and chaotic maps.This mechanism not only addresses the issue but also allows for an increased number of key spaces.(3).Potential pixel value alterations during encryption,especially during digit extraction or concatenation,lead to a vulnerability that can decrease encryption quality.We have implemented a mechanism that utilizes homomorphic encryption to understand image characteristics,enhancing visual accuracy,mitigating risks,and ensuring data integrity.