Eastman-kodak Company

Sumito Yoshikawa

Kazunobu Takahashi

Keiichi Mori

Kenichi Nakajima

Hidehiko Sato

To enhance accuracy in correction of a defective pixel. An image of an object is captured using a digital camera, and stored in an image memory. A processordetects a defect, using pixels located in the vertical and horizontal directions relative to, in lines of the same color as, and separated by one line from, the focused line. With any defect found, the presence or absence of line crawl is next determined. When it is determined that line crawl is present, the presence or absence of a diagonal edge is then determined. Subsequently, the surrounding pixels used to correct a defective pixel are selected in consideration of whether or not line crawl or a diagonal edge are present.

Pixel defect correction device for line crawl

A more preferable WB gain is calculated during strobe illuminated photography. When a WB gain for an image captured through strobe illuminated photography is calculated, the degree of contribution of strobe light is first estimated. When the degree of contribution is estimated, the image is first divided into a plurality of blocks, and typical brightness values and typical color values of the respective blocks are determined (S). Subsequently, the highest-brightness block among the blocks analogous to the color of strobe light is identified as a strobe color block (S). When there is a strobe color block, the degree of contribution of strobe light is estimated on the basis of the position of the strobe color block, a difference between the brightness of the strobe color block and the brightnesses of other blocks, and the like (Sto S). Meanwhile, when there is no strobe color block, the degree of contribution is estimated on the basis of the position of the highest-brightness block having the highest typical brightness value among the plurality of blocks (S, S).

Gain calculating device

The present invention provides a digital camera that is capable of capturing a picture in which the brightness between a main subject and a nighttime background are well balanced. When an image capture instruction is received while a night scene portrait mode is set, in a pixel addition output mode, an image sensoris set to a high sensitivity state by performing pixel addition, and outputs low resolution non-flash image data. Further, in an overall pixel output mode, the image sensoroutputs high resolution flash image data in a low sensitivity state without performing pixel addition. Next, a resolution conversion circuitconverts a resolution of non-flash image data to be equivalent to a resolution of flash image data. Then, an image combining circuitadds respective pixel values of corresponding pixels in the non-flash image data and the flash image data having the same resolution and combines both sets of image data to generate image data for recording.

Digital camera, image capture method, and image capture control program

To provide an auto white balance apparatus capable of performing stable white balance adjustment. A parent block is divided into child blocks by a block-dividing circuitand a typical value is calculated every child block by a typical-value calculating circuitand then the distance between the typical color difference component of each parent block and the typical color difference component of each child block is obtained on a color difference plane. A child block satisfying a condition equal to or less than a threshold value S is selected to recalculate reliability of a parent blocks by using only selected child blocks. Moreover, a white balance evaluating circuitcorrects reliabilities of parent blocks in accordance with the number of unselected child blocks.

Auto white balance apparatus and white balance adjusting method

Provide an electronic still camera using a color filter array (CFA) and a single CCD making it possible to simultaneously execute the processing for color signal interpolation and the processing for resizing. In the electronic still camera according to the present invention, an R signal, a G signal, and a B signal are output from each pixel of an image sensorhaving a CFA. A processorsimultaneously executes processing for interpolation and processing for resizing by computing (R, G, B) values at a given pixel position according to the color signals. Low frequency components R, G, Bof the color signal at the given pixel position are computed from signals from adjoining pixels, and the high frequency component Sis computed from a brightness value Y. The image data interpolated and changed to a desired size is stored in a memory, or is output via an interface I/Fto an external device such as a computer or a printer.

Growing community of inventors

Kanagawa, Japan

Sumito Yoshikawa