Event cameras are bio-inspired sensors that capture per-pixel brightness changes asynchronously, outputting a stream of events that encode the time, location, and sign of these changes. They offer high temporal resolution (μs), very high dynamic range (140 dB), low power consumption, and high pixel bandwidth, making them suitable for robotics and computer vision in challenging scenarios. However, novel methods are needed to process their unconventional output. This paper provides a comprehensive overview of event-based vision, including the working principle, available sensors, and applications from low-level to high-level vision tasks. It discusses techniques for processing events, such as learning-based methods and specialized processors like spiking neural networks. The paper also highlights remaining challenges and future opportunities in the field.Event cameras are bio-inspired sensors that capture per-pixel brightness changes asynchronously, outputting a stream of events that encode the time, location, and sign of these changes. They offer high temporal resolution (μs), very high dynamic range (140 dB), low power consumption, and high pixel bandwidth, making them suitable for robotics and computer vision in challenging scenarios. However, novel methods are needed to process their unconventional output. This paper provides a comprehensive overview of event-based vision, including the working principle, available sensors, and applications from low-level to high-level vision tasks. It discusses techniques for processing events, such as learning-based methods and specialized processors like spiking neural networks. The paper also highlights remaining challenges and future opportunities in the field.