Signal processing device

A signal processing device comprising:

• an imaging element in which an imaging region is divided into two left and right regions, and an electric signal is read for each divided region;
• synthesis means for synthesizing the electric signal read from the imaging element to generate signals for one screen;
• smoothing means for performing a low pass filter process in the vertical direction on the signal of each pixel of a gain deriving region near the boundary of the divided regions in one screen generated by the synthesis means;
• gain deriving means for deriving a gain for one divided region defined in advance for each horizontal line in the gain deriving region based on the signal after filter processing of the gain deriving region obtained by the smoothing means;
• average value deriving means for deriving an average value of the gain derived for each horizontal line in the gain deriving region by the gain deriving means;
• and
• correcting means for correcting a signal level difference between the both divided regions by multiplying the gain average value derived by the average value deriving means to the signal in the one divided region defined in advance out of the signals for one screen output from the synthesis means;
• wherein
• the gain deriving means derives the gain by deriving, from the signals of two same filter pixels of the one divided region, a predicted value of the signal of one same filter pixel of the other divided region based on the signals of a plurality of same color filter pixels near the boundary of the divided regions for each horizontal line in the gain deriving region, and comparing the derived predicted value and an actual signal value of the relevant pixel.

A signal processing device comprising:

• an imaging element in which an imaging region is divided into two left and right regions, and an electric signal is read for each divided region;
• synthesis means for synthesizing the electric signal read from the imaging element to generate signals for one screen;
• gain deriving means for deriving a gain for one divided region defined in advance for each horizontal line in one screen based on the signals of a plurality of pixels near the boundary of the divided regions out of the signals for one screen generated by the synthesis means;
• first average value deriving means for deriving an average value of the gain derived for each horizontal line in one screen by the gain deriving means;
• second average value deriving means for deriving an average value of the gain for the horizontal line in which a signal value of the pixel near the boundary of the divided region is less than or equal to a first predetermined value out of the gains derived for each horizontal line in one screen by the gain deriving means;
• and
• correcting means for correcting a signal level difference between the both divided regions by, out of the signals in one divided region defined in advance in the signals for one screen output from the synthesis means, multiplying the gain average value obtained by the second average value deriving means for the signal having the signal value of less than or equal to a second predetermined value, multiplying the gain average value obtained by the first average value deriving means for the signal having the signal value of greater than or equal to a third predetermined value greater than the second predetermined value, and multiplying a value obtained by weighing and adding the gain average value obtained by the first average value deriving means and the gain average value obtained by the second average deriving means according to the relevant signal value for the signal having the signal value between the second predetermined value and the third predetermined value;
• wherein
• the gain deriving means derives the gain by deriving, from the signals of two same filter pixels of one divided region, a predicted value of the signal of one same filter pixel of the other divided region based on the signals of a plurality of same color filter pixels near the boundary of the divided regions for each horizontal line in one screen, and comparing the derived predicted value and an actual signal value of the relevant pixel.

A signal processing device comprising:

• an imaging element in which an imaging region is divided into two left and right regions, and an electric signal is read for each divided region;
• synthesis means for synthesizing the electric signal read from the imaging element to generate signals for one screen;
• smoothing means for performing a low pass filter process in the vertical direction on the signal of each pixel of a gain deriving region near the boundary of the divided regions in one screen generated by the synthesis means;
• gain deriving means for deriving a gain for one divided region defined in advance for each horizontal line in the gain deriving region based on the signal after filter processing of the gain deriving region obtained by the smoothing means;
• first average value deriving means for deriving an average value of the gain derived for each horizontal line in the gain deriving region by the gain deriving means;
• second average value deriving means for deriving an average value of the gain of the horizontal line in which a signal value of the pixel near the boundary of the divided regions is less than or equal to a first predetermined value out of the gains derived for each horizontal line in the gain deriving region by the gain deriving means;
• and
• correcting means for correcting a signal level difference between the both divided regions by, out of the signals in one divided region defined in advance in the signals for one screen output from the synthesis means, multiplying the gain average value derived by the second average value deriving means for the signal having the signal value of less than or equal to a second predetermined value, multiplying the gain average value obtained by the first average value deriving means for the signal having the signal value of greater than or equal to a third predetermined value greater than the second predetermined value, and multiplying a value obtained by weighing and adding the gain average value obtained by the first average value deriving means and the gain average value obtained by the second average deriving means according to the relevant signal value for the signal having the signal value between the second predetermined value and the third predetermined value;
• wherein
• the gain deriving means derives the gain by deriving from the signals of two same filter pixels of one divided region, a predicted value of the signal of one same filter pixel of the other divided region based on the signals of a plurality of same color filter pixels near the boundary of the divided regions for each horizontal line in the gain deriving region, and comparing the derived predicted value and an actual signal value of the relevant pixel.

A signal processing device comprising:

• an imaging element in which an imaging region is divided into a plurality of regions in the lateral direction, and an electric signal is read for each divided region;
• synthesis means for synthesizing the electric signal read from the imaging element to generate signals for one screen;
• first gain deriving means for deriving, for every group consisting of two adjacent divided regions, a gain for one divided region defined in advance of the two divided regions in the group based on the signals for one screen generated by the synthesis means;
• second gain deriving means for deriving, with one divided region defined in advance of all the divided regions as a reference divided region, the gain for each of the other divided regions with respect to the reference divided region based on the gain for each group derived by the first gain deriving means;
• and
• correcting means for correcting a signal level difference between a plurality of divided regions by correcting the signal of the corresponding divided region out of the signals for one screen output from the synthesis means based on the gain for each of the other divided regions with respect to the reference divided region derived by the second gain deriving means;
• wherein
• the first gain deriving means includes,
• first means for deriving, for each group, the gain for one divided region defined in advance of the two divided regions of the group for each horizontal line in one screen based on the signals of a plurality of pixels near the boundary of two divided regions of the group out of the signals for one screen generated by the synthesis means, and
• second means for deriving, for each group, the gain for each group by deriving the average value of the gain for each horizontal line in one screen derived by the first means;
• in which
• the first means derives, for each group, the gain by deriving, from the signals of two same filter pixels of one divided region, a predicted value of the signal of the one same filter pixel of the other divided region based on the signals of a plurality of the same color filter pixels near the boundary of the two divided regions within the group for each horizontal line in one screen, and comparing the derived predicted value and an actual signal value of the relevant pixel.

A signal processing device comprising:

• an imaging element in which an imaging region is divided into a plurality of regions in the lateral direction, and an electric signal is read for each divided region;
• synthesis means for synthesizing the electric signal read from the imaging element to generate signals for one screen;
• first gain deriving means for deriving, for every group consisting of two adjacent divided regions, a gain for one divided region defined in advance of the two divided regions in the group based on a signals for one screen generated by the synthesis means;
• second gain deriving means for deriving, with one divided region defined in advance of all the divided regions as a reference divided region, the gain for each of the other divided regions with respect to the reference divided region based on the gain for each group derived by the first gain deriving means;
• and
• correcting means for correcting a signal level difference between a plurality of divided regions by correcting the signal of the corresponding divided region out of the signals for one screen output from the synthesis means based on the gain for each of the other divided regions with respect to the reference divided region derived by the second gain deriving means;
• wherein
• the first gain deriving means includes,
• first means for performing, for each group, a low pass filter process in the vertical direction on the signal of each pixel of the gain deriving region near the boundary of the two divided regions in the group out of the signals for one screen generated by the synthesis means,
• second means for deriving, for each group, the gain for one divided region defined in advance of the two divided regions in the group for each horizontal line in the gain deriving region of the group based on the signal after filter processing of the gain deriving region of the group obtained by the first means, and
• third means for deriving, for each group, the gain for each group by deriving the average value of the gain for each horizontal line derived by the second means;
• in which
• the second means derives, for each group, the gain by deriving, from the signals of two same filer pixels of one divided region, the predicted value of the signal of one same filter pixel of the other region based on the signals of a plurality of same color filter pixels near the boundary of the two divided regions within the group for each horizontal line in the gain deriving region of the group, and comparing the derived predicted value and an actual signal value of the pixel.