JPS63141453A - Interpolating method for defect picture - Google Patents

Abstract

PURPOSE:To improve the yield of a sensor while improving the picture quality and color reproducibility by correcting a defect picture element while replacing it by an output signal of a color filter nearest the picture element including the defective picture element. CONSTITUTION:In case of a defect of picture element Wj in a picture element (j) of channel A, although two data exist as a picture element Wi in the adjacent picture element (i) and a picture element Wk in a picture element (h) as a data to correct the defect picture element Wj, the defect is corrected by using the data of the picture element Wk nearest the defect picture element (j). If a detect picture element exists in the picture element Cj in the picture element (j) of a channel B, as the data to correct the picture element, the picture element Ci of the picture element (i) and the picture element Ck of the picture element (k) are to be considered, the defect is corrected by the data of the picture element Ci closest to the defect picture element (j). Since the defect picture element is corrected by using the picture element having the same color filter closest to the picture element having a defect picture element, the picture with color reproducibility is obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for supplementing defective pixels in a color linear image sensor for reading color originals.

[Conventional technology]

Conventionally, various image scanners have been developed as input terminal devices for reading originals into facsimiles, copying machines, computers, and the like. These scanners include
A so-called linear sensor in which many pixels are arranged one-dimensionally is generally used.

Furthermore, so-called area sensors in which multiple pixels are arranged two-dimensionally are used in television cameras and the like. In any case, the sensors used in these image scanners have a function of converting an optical density pattern on the surface of a document into an electrical density pattern.

Usually, if there are defective pixels in these sensors,
As an image reader, this causes a significant deterioration in image quality, and correction of defective pixels is a necessary condition. For example, in the case of the area sensor mentioned above, even in monochrome images and color images, it is possible to refer to the line before the line containing the defective pixel, and use the information of normal pixels vertically or diagonally above the defective pixel as data, and compare it. It is possible to obtain images with less deterioration in image quality. Instead, in the case of a linear sensor, it is known that in the case of a monochrome image, defective pixels are corrected by correlation between adjacent images. Furthermore, in the case of a color image, processing in the same manner as for a monochrome image is described in detail in, for example, Japanese Patent Publication No. 59-19513.

FIG. 3 is a plan view showing the color sensor disclosed in Japanese Patent Publication No. 59-19513 mentioned above. In the figure, h + 1 t jr k r L represents each picture element, and each of these picture elements hr 1 r jr k Ht is a pixel R with a red filter, a pixel G with a green filter, 3 consisting of pixel B with a blue filter
Each is made up of pixels.

In such a pixel configuration, for example, if pixel Gj becomes defective, that defect is replaced with the output signal value of pixel Gi three pixels before, and similarly, if another pixel Rj is defective, three pixels The defective portion is complemented by replacing the pixel with the output signal value of the previous pixel Ri, or, if the pixel row is defective, with the output signal value of the pixel Bi three pixels before.

[Problems to be Solved by the Invention] However, with such a pixel configuration, for example, in pixels G other than pixel B, information is shifted by about half a pixel from pixel i, and in pixel R, information is shifted from pixel i by about half a pixel. Approximately 5/6 pictures! There is a problem in that the color deviation from the original color of picture element j becomes large unless each of them is replaced with the shifted information.

The present invention has been made in view of the above-mentioned conventional problems, and its purpose is to improve image quality and color reproducibility by removing pixel misalignment, and to supplement defective pixels so that a color sensor can be obtained at low cost. It is about providing law.

[Means for solving problems]

The defective pixel complementation method according to the present invention includes a defective pixel including a defective pixel.
Among the picture elements before and after the picture element closest to the picture element, the output signal value of the pixel having the same color filter as that of the defective image is used to complement the output signal value of the defective pixel.

[For production]

The defective pixel complementation method in the present invention includes the defective pixel among the normal pixels before and after the defective pixel (two pixels before and after) that have the same color filter as the defective pixel. Since the output signal value from the pixel closest to the picture element is supplemented as the output signal value of the defective pixel, an image with good image quality and good color reproducibility can be obtained.

〔Example〕

Embodiments of the present invention will be described in detail below with reference to the drawings.

FIG. 1 is a plan view showing a pixel arrangement of a linear color image sensor, illustrating a defective pixel complementation method according to an embodiment of the present invention. In the figure, 1+1, t indicates each pixel constituting the linear color image sensor, and each pixel 1 r J , k rt is a transparent sensor pixel W without a filter, and a transparent sensor pixel W with a yellow filter. sensor pixel Y
, a sensor pixel G with a green filter, and a sensor pixel C with a cyan filter. Here, among these sensor pixels, pixel Yh* Wi p Y it Wj y Y
The horizontal direction (X-axis direction) arrangement of j r Wk... is the A channel, and the pixels Ch, Gi r Ci +
The horizontal direction (X-axis direction) arrangement of Gj r Cj . . . is assumed to be a B channel.

FIGS. 2(a) and 2(b) are plan views of a linear color image sensor illustrating a method for correcting pixel defects in the pixel array shown in FIG. 1. First, if the pixel Wj in the picture element i of channel A is defective as shown in FIG. Pixel Wi in pixel i and pixel Wk in the picture element
Although there are two pieces of data, correction is made using the data of the picture element Wk that is closest to the defective picture element j. This is the same correction rule when, for example, pixel Gj of the B channel is defective. In addition, as shown in the same figure (b), the pixel defect is B
Assuming that there is a pixel Cj in the picture element i of the channel, there are two data, the pixel Ci in the picture element i and the pixel ck in the picture element, as data for correcting the pixel Cj in the picture element i. , the defective pixel j is corrected using the data of the pixel Ci closest to it. This is the same rule as when the A channel pixel Yj is defective.

By correcting a defective pixel using the same pixel data of the nearest pixel according to such correction rules, image quality can be significantly improved and an image with excellent color reproducibility can be read.

In addition, if there are 2 or more defective pixels and 4 or less defective pixels in one picture element, the defect correction rule is basically the same as the one correction rule described above, and the defect is Inside 1
Even if the number varies between 1 and 4, it can be corrected using the rules shown in the table below.

In the above-mentioned embodiment, the picture elements are two pixels in the vertical direction,
It has a matrix shape of 2 pixels in the horizontal direction, and the A channel is a color filter array of pixels W and Y, and the B channel is a color filter array of pixels G,
Although a case has been described in which a color filter array of C is used, the present invention is not limited to this. The A channel is a color filter array of pixels W and G, and the B channel is a color filter array of pixels Y,
A picture element consisting of a combination of color filter arrays of C, A channel is a pixel WICO color filter array, B channel is a picture element consisting of a combination of color filter arrays of pixels G and Y, and A channel is a picture element iW, a color filter array of Y. , B
The channel may be a picture element made up of a combination of color filter arrays of pixels C and G, but it goes without saying that the same rule can be applied to both of them, and the same effect can be obtained. Note that in any case, it is clear that this correction rule cannot be used if the pixel to be used for correction is defective.

〔Effect of the invention〕

As explained above, according to the present invention, the defective pixel is compensated for by replacing the defective pixel with an output signal value having the same color filter closest to the pixel containing the defective pixel, thereby improving image quality and Extremely excellent effects such as improved color reproducibility, improved sensor yield, and a reduction in sensor prices can be obtained.

[Brief explanation of the drawing]

FIG. 1 is a plan view showing the pixel arrangement of a color image sensor according to the present invention, and FIGS. 2(a) and (b) are pixel arrays of the color image sensor for explaining the defective pixel complementation method according to the present invention. FIG. 3 is a plan view showing a conventional color sensor. 1itj+, t... picture element, Wi, Wj. Wk, Wt...Pixel without filter, Yh. Y i y Y j e Yk... Pixel with yellow filter, Gi, Gj, Gk, Gt Il...
・Pixels with green filter, Ch, C1, Cj, Ck
...Pixel with cyan color filter, A...-A
Channel, B111l・-B channel.

Claims (1)

[Claims] 1 consisting of 2 pixels arranged vertically and 2 pixels arranged horizontally.
A plurality of picture elements are arranged in the vertical and horizontal directions, and each picture element is transparent,
In a color linear sensor configured with a sensor covered with yellow, green, and cyan color filters, a defective pixel of the picture element is determined by the output signal value of the pixel of the same color filter in the picture element closest to the picture element containing the defective pixel. A defective pixel complementation method characterized by replacing and complementing with.