JPS63114253A - Linear image sensor - Google Patents

Abstract

PURPOSE:To remove noises perfectly and to improve the quality of an image, by providing black reference picture elements at gap parts between light sensitive picture elements, sequentially reading the signals generated in the light sensitive picture elements and the signals generated in the black reference picture elements, and correcting charge at dark time. CONSTITUTION:Light sensitive picture elements 21 are arranged at a specified pitch. Black reference picture elements 22 are arranged at gap parts between the light sensitive picture elements as dummy picture elements. A reading register 24 is arranged in parallel with the line of the light sensitive picture elements 21 and the black reference picture elements 22. An output circuit 25 is provided at one end of the register 24. A shift gate 23 is provided between the light sensitive picture elements 21 and the black reference picture element 23 in order to transfer the charge generated in each picture element. The charge at dark time is read out of the black reference picture elements 22 provided at the gap parts between the light sensitive picture elements at any time. The signal charge is corrected by using the charge at the dark time. Thus the correction with the charge at the dark time can be performed without the effect of temperature distribution and the like in a linear image sensor, and the excellent picture elements can be obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) The present invention relates to a linear image sensor, and particularly relates to a linear image sensor in which pixel lengths and pitches between pixels are different.

(Prior Art) There are many linear image sensors in which the length of a pixel and the pitch between pixels are not the same. For example, FIG. 3 shows a sensor in which these have a 2:1 relationship. According to the figure, photosensitive pixels 2 that generate and accumulate signal charges according to incident light are arranged at a predetermined pitch, and black reference pixels 2 that generate and accumulate dark charges are arranged on the extension of this photosensitive pixel row. 1 is formed. A readout register 4 is provided in parallel with the photosensitive pixel array to take out the charges generated from these pixels, and an output circuit 5 is provided at the tip thereof to output signal charges in a predetermined format. A shift gate 3 is provided between the black reference pixel 1 and the photosensitive pixel 2 and the readout register 4. The shift gate 3 is opened at a predetermined timing to transfer the signal charges accumulated in the photosensitive pixels all at once to the readout register.

Further, the read register 4 is composed of, for example, a CCD, and
The contents of the registers are sequentially transferred to the output section using clock signals such as phase and four-phase.

The output circuit 5 corrects the signal charge obtained by the black reference pixel 1 by subtracting the dark charge obtained by the black reference pixel 1 from the signal charge obtained by the photosensitive pixel 2, and outputs the signal charge at a predetermined level and timing. This type of correction is performed because photosensitive pixels always have a dark charge (dark current), which is a noise component that occurs even when no light is incident, so it is necessary to remove this to improve image quality. This is because there is.

In a linear image sensor with such a configuration, only one black reference pixel is provided in one chip, so if there is variation in dark charge across the entire linear image sensor, correction will not be performed sufficiently. It turns out. For example, heat distribution becomes irregular due to differences in heat generation and heat dissipation within the sensor, and dark charges increase in high temperature areas. Further, the amount of dark charge varies depending on the state of the manufacturing process and the output variation of the drive circuit.

If the dark charge is not sufficiently corrected, the image quality will deteriorate.

(Problems to be Solved by the Invention) As described above, in the conventional linear image sensor, it has been difficult to obtain good image quality because variations in dark charge due to temperature distribution in the sensor cannot be sufficiently corrected.

The present invention is an attempt to solve such problems and is capable of obtaining good image quality.

[Structure of the invention]

(Means for Solving the Problems) According to the present invention, photosensitive pixels that generate and accumulate signal charges according to incident light are arranged in a row at a predetermined pitch;
A black base pixel, which is provided in the gap between the photosensitive pixels and generates and accumulates dark charge regardless of the presence or absence of incident light, and a readout register that transfers the charge generated in the photosensitive pixel and the black reference pixel to the output section. This reading register sequentially reads out the signal charge in the photosensitive pixel and the dark charge in the black reference pixel.

(Function) In the linear image sensor according to the present invention, the dark charge is read out at any time from the black reference pixel provided in the gap between the photosensitive pixels, and this is used to correct the signal charge. . Therefore, it is possible to perform more complete dark charge correction without being affected by temperature distribution, etc. within the linear image sensor, and it is possible to obtain good image quality.

(Embodiment) Hereinafter, one embodiment of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a plan view showing an embodiment of the present invention, in which photosensitive pixels 1 are arranged at a predetermined pitch, and black reference pixels 22 are arranged as so-called dummy pixels in the gaps between the photosensitive pixels. It is set up. A readout register 24 is arranged parallel to the columns of the photosensitive pixels 21 and the black reference pixels 22, and an output circuit 25 is provided at one end thereof. Photosensitive pixel 21, black reference pixel 22, and readout register 23
A shift gate 23 is provided between them for transferring charges generated in each pixel. In this embodiment, the lengths of the photosensitive pixel 21 and the black reference pixel 22 are approximately equal, but this ratio may be applied as is to a linear image sensor having an arbitrary ratio of pixel length/pixel pitch. Note that when a black reference pixel is sandwiched between photosensitive pixels in this way, the distance between the photosensitive pixels may become somewhat longer, but this is not a problem at all if the user's use of the sensor does not require very high precision.

FIG. 2 is a graph showing the output potential when a certain pattern is read using the linear image sensor of FIG. According to the figure, corresponding to the black reference pixel 21, signals 31□, 31°, . . . due to dark charges and signals 32, 32□, . Note that a signal corresponding to the dark charge shown by hatching is superimposed on the signal 32 obtained from the photosensitive pixel 22, and for example, the signal 32. By subtracting the adjacent signal 341 from the signal 341, the dark charge can be almost completely corrected. Although such subtraction between signals is performed in the output circuit 25, it may also be performed in an external circuit.

In the above embodiment, a black reference pixel is provided in all the gaps between photosensitive pixels, but depending on the required resolution and reading accuracy, one black reference pixel may be provided in several photosensitive pixels.
Black reference pixels may be provided at a ratio of .

〔Effect of the invention〕

As described in detail based on the above embodiments, according to the present invention, a black reference pixel is provided in the gap between the photosensitive pixels, and a signal generated by the photosensitive pixel and a signal generated by the black reference pixel are sequentially read out. Since the dark charge is corrected using the image sensor, noise can be completely removed and the image quality can be improved.

[Brief explanation of the drawing]

FIG. 1 is a plan view showing an embodiment of a linear image sensor according to the present invention, FIG. 2 is a graph showing a signal waveform when reading out a pattern using the sensor shown in FIG. 1, and FIG.
The figure is a plan view showing an example of a conventional linear image sensor. 1.21...Black reference pixel, 2.22...Photosensitive pixel,
3.23...Shift gate, 4,24...Read register, 5.25...Output circuit.

Claims (1)

[Claims] 1. Photosensitive pixels arranged in a row at a predetermined pitch and generating and accumulating signal charges according to incident light; and photosensitive pixels arranged in a gap between the photosensitive pixels and depending on the presence or absence of incident light a black reference pixel that generates and accumulates dark charge regardless of the time; a readout register that transfers the charges generated in the photosensitive pixel and the black reference pixel to an output section; and a readout register that transfers the signal charge in the photosensitive pixel and the black reference A linear image sensor that corrects dark charges by sequentially reading out dark charges in pixels. 2. The linear image sensor according to claim 1, wherein the output section corrects the signal charge based on the fluctuation of the detected dark charge. 3. The linear image sensor according to claim 1, wherein a black reference pixel is provided in each gap between the photosensitive pixels.