<![CDATA[The need for a system capable of accurately and rapidly verifying academic documents is crucial for ensuring the authenticity of digital certificates. This research implements a digital signature mechanism within the ECert Verify (Electronic Certificate Verification) system, utilizing the RSA-2048 algorithm alongside the PSS (Probabilistic Signature Scheme) padding scheme and the SHA-256 hash function to authenticate digital certificate images from Coursera. Each certificate is automatically converted into a 300 DPI (3508×2480 pixel) grayscale image to maintain hash consistency. The system was developed using Python with the pyca/cryptography library, Django for the backend, and React (Vite) for the frontend. Testing involved 20 datasets, comprising 5 authentic certificates and 15 manipulated samples (involving screenshots, re-compression, and content modification). Test results demonstrated a 100% verification accuracy rate, with no instances of false positives or false negatives. An average Avalanche Effect of 50.88% indicates that SHA-256 is highly sensitive to minor changes in the certificate image. Shannon Entropy testing yielded an average value of 7.1786 bits (89.73%)—close to the theoretical maximum of 8.0 bits—demonstrating a high level of randomness in the generated digital signatures. The average verification time was just 2.46 ms per file, indicating strong performance in a controlled single-user testing environment. Furthermore, a limited user acceptance survey regarding the interface's operational aspects revealed that the system is highly rated and functional in supporting the certificate issuance process for operators.]]>
