A computer vision algorithm that detects objects in view of a camera or other sensor: for example pedestrians, vehicles, animals, or cyclists.
Algorithms are typically artificial intelligence programs that have been trained on large datasets (i.e. 1000 examples of a Stop sign) in order that variances in the driving environment, including dirty or damaged signs, can still be interpreted correctly.
Sensor data from one source (such as a camera) may be validated against a second source (map data or LiDAR) in order to ensure the data validity.
