Investigators are using new technology to improve how crime scenes are analyzed, making the process faster and more effective. This study used a mobile phone with EpocCam Pro to record video in real time at the crime scene, which is processed on a laptop with Elgato Camera Hub. The system uses YOLOv8n for object detection and ResNet-18 for classification. When an object is detected, it is cropped, labeled, and stored in organized folders. This allows investigators to access and manage evidence as needed quickly. Our system worked effectively by achieving 87% accuracy in object detection (mAP@0.5) and 92.3% accuracy in classification. It consistently operated at 25 FPS, enabling real-time processing with minimal lag. Incorporating mobile technology with AI, this study improves forensic investigations by streamlining how evidence is identified, categorized, and stored.