| 摘要 |
<P>PROBLEM TO BE SOLVED: To compress a RAW image without generating unnatural gray scales. <P>SOLUTION: The number x of photons is converted into voltage and converted into a digital value a. Next, a digital signal a is temporarily converted into the number x of electrons generated on an image sensor. Then, the number x of electrons is converted into a non-linear value y=f(x) by using a signal conversion formula using a quantization step function r(x) before noise removal calculated in advance. Since the quantization step function r(x) before noise removal is determined so that the difference between the expectation value of a signal before quantization and the expectation value of a signal after quantization becomes smaller, a RAW image signal can be compressed to have a fewer number of gray scales without generating unnatural gray scales by the above-described signal conversion processing. <P>COPYRIGHT: (C)2012,JPO&INPIT |
