Canon Says the Future of Cameras Is AI

The next major revolution in cameras may not be about more megapixels. It could be artificial intelligence running directly inside the camera. Canon has outlined how it sees AI becoming a larger part of future photography, but its approach is different from simply using generative AI to create things that were never in the original scene. Tetsushi Hibi, advisory director of Canon’s Imaging Optical Business Unit, explained that Canon wants AI to help cameras reproduce scenes as closely as possible to the reality of the moment being photographed. The company points to applications including noise reduction, color correction, distortion correction and optical-aberration correction. And some of that future is already here. Canon’s EOS R1 and EOS R5 Mark II use deep-learning technology for advanced subject recognition and tracking. The processing system can analyze people, objects and actions in real time to help the autofocus system stay locked onto the important subject. AI is also moving deeper into image processing. The EOS R1 can perform deep-learning-based noise reduction inside the camera and use neural-network upscaling to double an image’s horizontal and vertical pixel count while attempting to preserve realistic detail. The EOS R5 Mark II also includes Neural Network Noise Reduction, using deep learning to reduce high-ISO noise directly inside the camera. This points toward a much larger shift in photography. Professional cameras are increasingly able to understand what they are looking at, identify the important parts of a scene and perform processing that once required a computer and specialized software. At the same time, consumer AI tools are simplifying the editing process itself. For example, if there is an unwanted cable or object in a photograph, tools such as Alice AI can let the user choose an object-removal function and let AI handle the technical editing. That changes the skill photographers may need in the future. Instead of knowing every masking tool, selection method and retouching technique, users may increasingly need to understand what they want the final image to look like — and communicate that clearly to AI. Canon’s approach also highlights an important distinction between AI-assisted photography and AI-generated imagery. One tries to help the photographer capture and process the real scene more effectively. The other can create visual information that never existed in front of the lens. As cameras become more intelligent, that distinction could become increasingly important. What will matter more for the photographer of the future: knowing how to use complex editing software, or knowing how to tell AI exactly what result you want? Subscribe to Xenon Lab News for more AI, photography and future technology. #Canon #CanonEOS #EOSR1 #EOSR5MarkII #AIPhotography #ArtificialIntelligence #CameraAI #Photography #FutureCameras #DeepLearning #ComputationalPhotography #PhotoEditing #AliceAI #CameraTechnology #FuturePhotography #TechNews #XenonLabNews