The application of electronic evidence in forensic imaging analysis

In the modern judicial system, especially in forensic image analysis, electronic evidence is becoming increasingly important. This article focuses on the application of advanced algorithms and technologies to improve the efficiency and accuracy of electronic evidence in this field. In this study, an image analysis framework based on deep learning is proposed, where key features of the images are automatically extracted using convolutional neural networks (CNN), and an enhancement algorithm is designed for blurred or low-resolution images. In the experiments, X-ray images, CT scans, and video clips were selected for testing. The results show that the algorithm improves the accuracy of image analysis, achieving a detection accuracy of 96.5 % in binary classification tasks and maintaining an accuracy of 84.7 % even under high noise conditions (SNR = 5 dB). Additionally, the processing time was significantly reduced, with a training time of only 3.2 h and an inference time of 45 ms per sample. The algorithm successfully identifies key evidence in actual criminal cases. After analysis, it is found that the algorithm has high robustness and stability in complex scenes. Future research work will explore more efficient algorithm models and extend them to related fields. Highlights • A DL-based image analysis framework is proposed for forensic images, automatically extracting key evidence features via CNN. • A new enhancement algorithm for blurred/low-res forensic images is proposed, improving quality and revealing hidden key evidence. • The algorithm proves efficient and robust, shortening processing time for key evidence identification in criminal cases.