JPH0271682A - Picture read element correction device - Google Patents

Abstract

PURPOSE:To execute the detection and correction of a defect picture element simultaneously and to cope with picture defects of many kinds over a wide range by using a filter, giving a prescribed luminous quantity, measuring the luminous quantity change of each picture element of an image sensor with respect to the prescribed luminous quantity, detecting a defective picture element, storing it and correcting it for each picture element. CONSTITUTION:The position of an ND filter 12 is adjusted to vary the luminous quantity into ND1-ND5 with respect to the image sensor 10 and whether or not an output DP of the sensor 10 is within a permissible range between set threshold levels L1 and L2 is checked by a detection circuit 20. A defect data DF for each picture element in response to the luminous quantity ND is stored in a RAM 4 via a gate 3. Then the filter 12 is removed to read a picture data of an original 14 by the sensor 10. In this case, the data DP from an A/D converter 2 is inputted to a correction circuit 30 and a data DPA relating to the defect from the RAM 4 is read and a picture clock PXC applies correction by interpolation of adjacent data and a picture data PS is outputted. Through the constitution above, even if the image sensor is replaced, simple correction is attained and the device processes various kinds of defects.

Description

[Detailed Description of the Invention] Purpose of the Invention: (Industrial Application Field) This invention automatically detects defective pixels in an image reading element of an image sensor shade made of CCD, MOS, etc. and corrects image data. The present invention relates to an image reading element correction device.

(Prior art) An image sensor made up of a CCU, etc. usually has variations in characteristics for each pixel, and conventionally, defective pixels in the image sensor have been detected for each pixel in advance in order to correct the variations in pixels of the image sensor. Then, which pixels are defective pixels are stored in memory in advance, and the image data is corrected when the image is actually read.

(Problems to be Solved by the Invention) The conventional defective pixel correction device described above cannot detect and correct defective pixels of an image sensor at the same time.
Defective pixels are detected and stored in advance in the image sensor, and image data is corrected in accordance with the defective pixels when reading an actual image. Therefore, it is not possible to detect and correct defective pixels at the same time, and it is not possible to perform correction for changes in the amount of light in each pixel.

This invention was made due to the circumstances mentioned above,
SUMMARY OF THE INVENTION An object of the present invention is to provide an image reading element correction device that can simultaneously detect defective pixels of an image reading element and correct the defective pixels.

Structure of the Invention; (Means for Solving the Problems) The present invention relates to a defective pixel detection and correction device for an image reading element. , pixel defect detection means for detecting pixel defects with respect to the amount of light varied by the light irradiation means for each pixel of the image reading element; storage means for storing defect data detected by the pixel defect detection means; This is achieved by providing pixel defect correction means for correcting the image read data of the image reading element for each pixel using the stored data from the storage means.

(Operation) In this invention, the light i? of the light source is applied to the image sensor. ■
Or, by applying a predetermined amount of light using a No (Neutral Density) filter and measuring whether the change in the scene of each pixel of the image sensor is within a predetermined range with respect to the predetermined amount of light, defects in the image sensor can be detected. It is designed to detect pixels. This defect detection data is stored in memory for each light intensity, and the image data is corrected at the defective pixel for the image data obtained by reading the actual image, so that defect detection and image data correction are performed simultaneously. image data can be output.

(Embodiment) FIG. 1 shows an embodiment of the present invention, in which the image output signal IP of the image sensor IO at a predetermined scene is inputted to the AD converter 2 via the sample and hold circuit 1, and converted into The digital image data DI' is input to the pixel defect detection circuit 20 and the pixel defect correction circuit 30, and the defect data DI' detected by the pixel defect detection circuit 20 is recorded in the IIAM 4 through the gate 3.

The memory data DP stored in 11AM4 is inputted to the pixel defect correction circuit 30 through the gate 3, and the image data DP obtained when reading the actual document is corrected and converted into image data PS.
Output as . Also, an image sensor lO. Sample hold circuit 1. The converter 21, the pixel defect detection circuit 20, and the pixel defect correction circuit 30 are controlled in synchronization with the pixel clock PxC, and the address counter 5 corresponds to the pixels of the image sensor 10 for data storage and data readout of the IIAM 4. I try to take it.

As shown in FIG. 2, light is irradiated onto the image sensor IO from a light source 11 via a lens 13, and between the lens 13 and the light source 11 there is a light source for adjusting the amount of light in a predetermined manner. For example, an NO filter 12 divided into five stages (NDI to N05) is provided, and by appropriately adjusting the position of the NO filter 12, a predetermined amount of light can be given to the image sensor IO. It has become. When reading the image data of the original 14, N
Place the O filter 12 outside the optical path and connect the original) 414 to the light source 1.
1 and the lens 13 for reading.

In such a configuration, the position of the NO filter 12 shown in FIG.
As shown in the figure, the pixel defect detection circuit 20 detects whether the output DP of the image sensor 10 when the light intensity is changed from NOI to ND5 is within the tolerance range between the set threshold values Ll and L2. do. This means that if there is no defect in the pixel of the image sensor IO, the pixel output value will be almost proportional to the increase in the amount of light.For example, the characteristic ① in Fig. 3 will all fall within the range of the threshold values Ll and L2. Since it is included, all light is 1NDl ~
It shows that there is no defect for N[lS, and the characteristic ■ causes a defect in light l No5 and other light 1ND
No defects exist in 1 to ND4, and the characteristic ■ is light ff1N.
It is shown that a defect occurs in DI and No. 2, but does not occur in light amounts ND3 to ND5. Such multi-light 1N
Defect data for each pixel corresponding to D1 to ND5 is obtained in the pixel defect detection circuit 20, and the defect data OF is
- stored in RAM 4 via port 3. Such defect detection of the image sensor IO is performed at the beginning of image reading, and the defect data DF is stored in the IIAM 4.

After that, the NO filter 12 is placed outside the optical path of the image sensor 10, and the original 14 is placed inside the optical path, and this image data is read by the image sensor 10. In this case, the image data DP from the converter 2 is is input to the pixel defect correction circuit 30, and the stored data DPA regarding the pixel defect is read out from the RAM 4 and inputted manually through the gate 3, and the data is input as shown in FIG. 6 (^) or (II) based on the pixel clock I'XC. The image data PS is output after correction by interpolation of adjacent data.

FIG. 4 shows an example of the pixel defect detection circuit 20, in which the window comparator 2! defines the above-mentioned allowable range to each light amount NDI-NO3@! The window comparators 21 to 25 that operate are configured with selection signals SLI to SL.
5 to be selected. Then, in the selected window comparator, binary defect data DFI to D are determined depending on whether or not the manually inputted image data OP is within the allowable range of the set threshold value.
It is designed to output F5. Therefore, regarding the characteristics ■ to ■ in Figure 3, the defect data DF as shown in Table 1 below is
I to DF5 are output.

Further, FIG. 5 shows a detailed configuration example of the pixel defect correction circuit 30, in which the pixel clock PxC is manually input to the counter 31 and latch circuits 33 to 35, and the counter 31 receives a counter clear signal input in synchronization with one line. It is cleared by TCL, and the output of counter 36 is IIAM.
/110M32, and the output of M/110M32 to R is input to the selector 37 as an A or B selection signal. Further, the image data DP is inputted to the latch circuit 33, and its output LTI is inputted to the latch circuit 34 and also inputted to the adder 36, and the output LT2 of the latch circuit 34 is inputted to the adder 36.
is input to the latch circuit 35 and the B of the selector 37
The output LT3 of the latch circuit 35 is input to the adder 36. Furthermore, the output data of the adder 36 is input to the A terminal of the selector 37, and the corrected image data PS is output from the selector 37. Note that the adder 36 is configured to shift the sum of data LTI and L12 by 1 bit (÷2) and output it, and when data LTI and L12 are abnormal values,
The output of the adder 36 is selected by the selector 37 and output as image data PS. Normally, the output LT2 of the latch circuit 34 is selected as normal data by the selector 37 and output as image data ps.

In the above description, the amount of light irradiated by the image sensor is changed to a predetermined position using an ND filter, but the amount of light emitted from the light source itself may be changed.

Effects of the Invention: As described above, according to the image reading element correction device of the present invention, the number of man-hours for adjusting image data can be reduced by m, and even if the image sensor is replaced, image data can be corrected on the cylinder 41. This method has the advantage of being able to deal with a wide range of types of pixel defects.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a diagram showing a means for changing the amount of light for an image sensor, FIG. 3 is a diagram showing an example of the characteristics of a pixel with respect to the amount of light, and FIG. 4 is a diagram showing a pixel Block configuration diagram showing an example of a defect detection circuit, No. 5
The figure is a block diagram showing an example of a pixel defect correction circuit, and FIGS. 6 and 7 are diagrams for explaining correction of pixel data. 1... Sample bold circuit, 2... An converter,
3... Gate, 4... 1^M, 5... Address counter, IO... Image sensor, 11... Light source,
12...NO filter, 13...lens, 14...
- Original document, 20... pixel defect detection circuit, 30... pixel defect correction circuit. , Figure JI5' times Lj5

Claims (1)

[Claims] 1. A light irradiation means for varying the amount of light irradiated to the image reading element, a pixel defect detection means for detecting a pixel defect with respect to the amount of light varied by the light irradiation means for each pixel of the image reading element, and this pixel defect detection and a pixel defect correction means for correcting the image read data of the image reading element for each pixel using the stored data from the storage means. Features: Image reading element correction device.