Exposure Control Device for Moving Objects

The patent segments the frame image into multiple regions based on feature quantities (such as edge intensity, gradient, or texture features). Pixels are divided into target regions (containing objects of interest) and non-target regions (such as sky or background). This segmentation allows selective use of pixels for exposure calculation, improving accuracy by excluding distracting regions while maintaining manageable complexity through automated region classification.

The patent applies different weighting or selection criteria to different regions of the image based on their feature quantities. Regions with high feature quantities (likely containing edges or important structures) are given higher priority for exposure calculation, while regions with low feature quantities are downweighted or excluded. This local quality approach ensures that exposure control is driven by visually significant areas rather than uniform treatment of all pixels.

The patent performs preliminary calculation of feature quantities for all pixels before the exposure control decision. By pre-computing feature quantities (such as edge detection or gradient magnitude) and storing them, the system avoids repeated calculations during exposure adjustment. This preliminary action separates the complex feature analysis from the real-time exposure control, improving target object exposure accuracy while managing computational complexity through staged processing.

The patent replaces complex manual or heuristic exposure adjustment mechanisms with an automated system based on feature quantity analysis. Instead of relying on simple luminance averaging or fixed-region sampling, the system uses computational feature detection (such as edge-based or gradient-based features) to automatically identify and weight relevant pixels. This substitution of mechanical/simple methods with intelligent algorithms improves target object exposure accuracy while the automation manages the inherent complexity.