Method and apparatus for detecting and filling-in dead spots of video signals from an image sensor

摘要

Dead spots within an array of photo-sensitive video signal generation elements are detected based on an excessive rate-of-change in sequentially-accessed pre-stored digital sensitivity correction coefficients corresponding to the array of photosensitive elements. The coefficients are sequentially accessed for use in making synchronous real-time sensitivity corrections to the non-dead video output from corresponding ones of the sensor elements. In response to such detected dead spots, an artificial video signal is also substituted during any dead spot so as to "fill-in" the "missing" portions of the video signal otherwise present due to dead spots in the array of video signal generating elements. In the exemplary embodiment, the leading and trailing edge of such dead spots is detected by comparing arithmetic differences between successive correction coefficients to a predetermined threshold value. The synthetic video signal to be substituted during the "fill-in" period may simply be a continuation of the last valid video signal value (e.g., as maintained by an enabled sample-and-hold circuit) or it may involve a local average of pixel values in the immediate vicinity of the pixel(s) being filled-in or a global average of pixel values taken from a major portion of the entire frame of video signals being processed.