Method and apparatus for correcting pixel image signal intensity
View Patent ↗When charges Q (x, y) transferred from an image inputting device 1 are to be converted into first signal intensities S′ (x, y) and signal processing is to be performed for the first signal intensity S′ (x, y) of a particular pixel, a maximum value Smax, a minimum value Smin and an average value Save are calculated from signal intensities S′ (x−1, y) and S′ (x+1, y) at adjacent pixels. When S′ (x, y)>Smax×A is satisfied, it is determined that the signal intensity S (x, y) at the particular pixel=Save×C (where A and C are coefficients), whereas when S′ (x, y)<Smin×B is satisfied, it is determined that the signal intensity S (x, y)=Save×D (where B and D are coefficients), and processing is performed so as to obtain an appropriate intensity S (x, y). This makes it possible to accurately judge a defective pixel attributable to dirt adhering to a pixel array of the image inputting device, a crystal defect, etc., and compensate a defect in an image caused by the defective pixel.
1. A signal processing method which is executed by a signal processing device for calculation of an image signal intensity by processing image signals from pixels sequentially output from an image inputting device in which plural pixels for photoelectric conversion of incident light are arranged in the horizontal direction and the vertical direction, said signal processing device executing:
a first step of calculating a maximum value, a minimum value and an average value of image signal intensities of plural pixels which are arranged around a certain pixel; and
a second step of executing correction of said image signal intensity of said certain pixel by replacing said image signal intensity with a value calculated by multiplying said average value by a coefficient in a range of 1 to 0.8 when said image signal intensity of said certain pixel is larger than a value which is calculated by multiplying said maximum value by a first coefficient, and executing correction of said image signal intensity of said certain pixel by replacing said image signal intensity with a value calculated by multiplying said average value by a coefficient in a range of 0.8 to 0.5 when said image signal intensity of said certain pixel is smaller than a value which is calculated by multiplying said minimum value by a second coefficient, for making said correction in a manner that said correction of said image signal intensity of said certain pixel relative to said image signal intensities of said plural pixels arranged around said certain pixel cannot be recognized effectively.
2. The signal processing method of claim 1 , wherein said image signal intensities are the sums of plural chrominance signal intensities.
3. An image capturing device, comprising an image inputting device in which plural pixels for photoelectric conversion of incident light are arranged in the horizontal direction and the vertical direction, a signal processing device which calculates an image signal intensity by processing image signals from pixels sequentially output from said image inputting device, and an image display device or an image storage device which receives said image signal intensity calculated by said signal processing device,
wherein said signal processing device is equipped with a function of calculating a maximum value, a minimum value and an average value of image signal intensities of plural pixels which are arranged around a certain pixel and correcting said image signal intensity of said certain pixel by replacing said image signal intensity with a value calculated by multiplying said average value by a coefficient in a range of 1 to 0.8 when said image signal intensity of said certain pixel is larger than a value which is calculated by multiplying said maximum value by a first coefficient, and correcting said image signal intensity of said certain pixel by replacing said image signal intensity with a value calculated by multiplying said average value by a coefficient in a range of 0.8 to 0.5 when said image signal intensity of said certain pixel is or smaller than a value which is calculated by multiplying said minimum value by a second coefficient, for making said correction in a manner that said correction of said image signal intensity of said certain pixel relative to said image signal intensities of said plural pixels arranged around said certain pixel cannot be recognized effectively.