JP2010034703A - Method and equipment for inspecting image sensor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an inspection method and equipment for suppressing an influence of disturbance caused by pixel defect and/or dust deposition and improving accuracy of inspection for output uniformity of an image sensor. <P>SOLUTION: Inspection image data are acquired from an image sensor. The acquired inspection image data are filtered. The inspection image data are divided into a plurality of blocks. An output average value is computed for each block. Based on difference between the maximum output average value and the minimum output average value of the plurality of blocks, an output uniformity is computed. Determination is made by comparing the output uniformity with a determination value set in advance. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image sensor inspection method and an inspection apparatus.

  Conventionally, an image pickup device such as a CCD image sensor or a CMOS image sensor is used in an image pickup apparatus such as a digital camera. The image sensor photoelectrically converts subject light with a photodiode and outputs a charge signal to generate image data. 2. Description of the Related Art An image pickup device is known that includes a color filter, and has a configuration capable of generating color image data by determining the color of light incident on each photodiode depending on the arrangement of the color filter.

  By the way, due to non-uniform optical characteristics of photodiodes, microlenses, color filters, etc. of imaging devices due to variations in manufacturing processes, non-uniform output values such as uneven output of image data and shading occur. There are things to do. Conventionally, in order to inspect the output uniformity, a means for obtaining an output level for each pixel by irradiating the image sensor with uniform light, or obtaining a difference amount of an average level of an area obtained by dividing image data at regular intervals Then, there is a means for determining uniformity based on the value of the difference amount.

Conventionally, as a method for inspecting output uniformity of an image sensor, for example, there is a method shown in the following patent document.
JP 2004-140714 A JP 2003-116042 A

By the way, a pixel defect may occur at the time of manufacture of the image pickup device, or dust may adhere to the surface of the image pickup region of the image pickup device. Then, when inspecting the image sensor, output values that are extremely far from the normal level are detected in areas where defective pixels or dust are attached, and output irregularities cannot be detected accurately, resulting in a decrease in inspection accuracy. Will end up.
In addition, output unevenness and shading due to non-uniform optical characteristics of the image sensor change gently over the entire image data, so it is necessary to measure the output value in the widest possible area. When the output value of each pixel is measured, the time required for the inspection is inevitably increased. In addition, since the pixel defect is also inspected as described above, the accuracy of the inspection cannot be changed.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide an inspection method and inspection capable of suppressing the influence of disturbance due to pixel defects, dust adhesion, etc. and improving the accuracy of the output uniformity inspection of the image sensor. To provide an apparatus.

The present invention is as follows.
(1) An image sensor inspection method for inspecting output uniformity of an image sensor,
Obtaining inspection image data from the image sensor;
Filtering the acquired inspection image data;
Dividing the inspection image data into a plurality of blocks, calculating an average output value of each block,
Calculating output uniformity based on the difference between the maximum output average value and the minimum output average value among the plurality of blocks;
An image sensor inspection method for performing pass / fail determination by comparing the output uniformity with a predetermined determination value.
(2) The inspection method according to (1) above,
When there is an abnormal block in which the output average value is abnormal among the plurality of blocks, the output average value of the abnormal block and a block around the abnormal block are compared, and the abnormal block is a singular point. A method for inspecting an image sensor that removes the abnormal block from an inspection target when it is determined that the abnormal block exists.
(3) An imaging device inspection apparatus for inspecting output uniformity of the imaging device,
Inspection image data acquisition means for acquiring inspection image data from the imaging element;
Filtering the acquired inspection image data;
Block average value calculating means for dividing the inspection image data into a plurality of blocks and calculating an average output value of each block;
Uniformity calculating means for calculating output uniformity based on a difference between a maximum output average value and a minimum output average value among the plurality of blocks;
An image sensor inspection device comprising: determination means for performing pass / fail determination by comparing the output uniformity with a predetermined determination value.
(4) The inspection apparatus according to (3) above,
When there is an abnormal block in which the output average value is abnormal among the plurality of blocks, the output average value of the abnormal block and a block around the abnormal block are compared, and the abnormal block is a singular point. An inspection device for an image sensor that removes the abnormal block from an inspection target when it is determined that the abnormal block exists.

  According to the present invention, the acquired inspection image data is filtered to remove the output value of the defective pixel, and then the inspection image data is divided into a plurality of blocks, and the output average value of each block is calculated. By doing so, it is possible to avoid detection of an output value that is extremely far from the normal level.

  In addition, when there is an abnormal block whose average output value is abnormal due to adhesion of dust etc. among multiple blocks, if it is determined that the abnormal block is a singular point, the abnormal block is subject to inspection. Therefore, only output unevenness and shading caused by non-uniform optical characteristics can be inspected with high accuracy.

  According to the present invention, it is possible to provide an inspection method and an inspection apparatus capable of suppressing the influence of disturbance due to pixel defects, dust adhesion, and the like and improving the accuracy of the output uniformity inspection of the image sensor.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a schematic plan view showing a configuration of a CCD image sensor (solid-state image sensor) as an example of an image sensor to be inspected. The imaging device to be inspected according to the present invention is not limited to a CCD image sensor, and may be a CMOS image sensor, for example.

  The solid-state imaging device 10 is provided for each of a plurality of photodiodes (PDs) 11 that convert incident light into signal charges corresponding to the amount of light and store it, and for each vertical column of the photodiodes 11. A plurality of vertical CCDs (VCCD) 12 for reading out charges and transferring them vertically, a horizontal CCD (HCCD) 13 for transferring signal charges transferred from the VCCD 12 horizontally, and a charge amount of signal charges transferred from the HCCD 13 And an output amplifier 14 that outputs a signal voltage (image data).

  The photodiode 11 has a substantially octagonal outer shape and is two-dimensionally arranged in a honeycomb shape at a predetermined pitch, and a color filter and a microlens (not shown) are stacked thereon. The color of light incident on the photodiode 11 is determined by a color filter. The photodiode 11 with R is red light, the photodiode 11 with G is green light, and the photodiode 11 with B is used. Blue light is incident on.

  The VCCD 12, the HCCD 13, and the output amplifier 14 are driven by drive pulses input from the outside of the solid-state imaging device 10, sequentially read out the signal charges accumulated in the photodiode 11, and transfer the image data after vertical and horizontal transfer. Outputs externally.

  FIG. 2 shows an inspection apparatus for inspecting the uniformity of the output value of the image data output from the solid-state imaging device. The inspection device 30 includes an arithmetic processing device 21, an analog signal processing unit 23, an A / D converter 24, a digital signal processing unit 25, a recording unit 26, and a display unit 27. The inspection device 30 is comprehensively controlled by the arithmetic processing unit 21. The CCD drive unit 22 generates a drive pulse group including a read pulse, a vertical transfer pulse, a horizontal transfer pulse, and the like to drive the solid-state image sensor 10 and cause the solid-state image sensor 10 to output image data. At the time of inspection, the CCD 10 is irradiated with uniform light from a light source. The inspection device 30 may be provided with a light source and a light source control unit (not shown) that controls the light source.

  The analog signal processing unit 23 receives analog signal image data output from the CCD 10, performs predetermined signal processing, and then outputs the image data to the A / D converter 24.

  The A / D converter 24 converts the image data output from the analog signal processing unit 23 into a digital signal and inputs the digital signal to the digital signal processing unit 25. The digital signal processing unit 25 performs interpolation processing (synchronization processing) on the input image data, and generates pixel data of a color that is insufficient for each pixel. For example, for R (red) pixels, G (green) and B (blue) pixel data are estimated from the surroundings.

  The digital signal processing unit 25 inputs image data having the same number of R, G, and B pixel data by this interpolation processing to a recording unit 26 configured by a memory or the like. The recording unit 26 stores the input image data.

  The arithmetic processing unit 21 includes an inspection image data acquisition unit 31, a white balance (WB) processing unit 32, a smoothing processing unit 33, a block average value calculation unit 34, a uniformity calculation unit 35, and a determination unit 36. Yes. The inspection image data acquisition unit 31 acquires the image data stored in the recording unit 26 as inspection image data when the image sensor is inspected.

  The WB processing unit 32 performs a predetermined evaluation for determining whether a white subject is reproduced as white, and performs gain adjustment of the inspection image data based on the obtained evaluation value.

  The smoothing processing unit 33 performs averaging processing, median processing, and the like for each predetermined area in the inspection image data in order to remove disturbances such as output values of defective pixels included in the inspection image data, and smoothes I do. When the averaging process is used, the output value of the defective pixel in the inspection image data is replaced with an average value calculated for the output value of the defective pixel and the output value of the pixels adjacent to the periphery of the defective pixel. When the median processing is used, the output value of the defective pixel is replaced with the median value calculated for the output value of the defective pixel and the output values of the neighboring pixels.

  The block average value calculation unit 34 divides the inspection image data into a plurality of blocks. FIG. 3 is a diagram schematically illustrating a state in which the inspection image data is divided into a plurality of blocks. As shown in FIG. 3, the inspection image data has n rows in the x direction when the horizontal direction in the figure is the x direction and the vertical direction is the y direction. It can be divided into m × n blocks of m rows in the y direction. At this time, both m and n are natural numbers or positive integers. Each block is composed of i × j pixels in i columns in the x direction and j rows in the y direction. The output average value of each block is equivalent to the average value of the output values of each pixel constituting the block.

  In the inspection image data shown in FIG. 3, for example, the output value of the block b1 closest to the image center is the largest, and the output value of each block b2 located at the four corners farthest from the image center is the smallest. ing. The inspection image data has different output average values for each block in the x direction and the y direction. In FIG. 3, the closer the image is to the center of the image, the larger the output value is, and the farther from the image center, the smaller the output value is. Show.

  The uniformity calculation unit 35 calculates the difference between the block having the maximum output average value and the block having the minimum output average value among the plurality of blocks as the output uniformity of the inspection image data. The smaller the output uniformity value, the closer the output value of the inspection image data becomes, and the larger the output uniformity value, the more the output value of the inspection image data varies.

  The determination unit 36 compares the output uniformity of the inspection image data calculated by the uniformity calculation unit 35 with a predetermined determination value. If the determination value exceeds the determination value, output unevenness, shading, or the like is included in the inspection image data. It is determined that has occurred, and a failure is determined. On the other hand, when the output uniformity is equal to or less than the determination value, it is determined that the output is non-defective product that does not generate output unevenness or shading in the inspection image data, and the pass is determined.

  The inspection device 30 displays the result of the pass / fail determination by the determination unit 36 on the display unit 27.

  Next, the procedure of the inspection method according to the present invention will be described. In the following description, the procedure for executing the inspection method using the inspection apparatus shown in FIG. 2 will be described as an example, and the description of the already described configuration and its operation will be omitted.

FIG. 4 is a flowchart showing the procedure of the inspection method.
First, the inspection image data acquisition unit 31 acquires inspection image data. The acquired inspection image data is subjected to white balance processing by the WB processing unit 32 and filtered by the smoothing processing unit 33. The output value of the defective pixel included in the inspection image data is removed by filtering.

  After the smoothing process, the inspection image data is divided into n × m blocks as shown in FIG. Then, an average output value is calculated for each block. By calculating the output average value for each block, the influence of disturbance can be smoothed. The number of blocks into which the inspection image data is divided is not particularly limited, and can be arbitrarily set as long as the influence of disturbance can be smoothed.

  Next, among the plurality of blocks, the block having the maximum output average value and the block having the minimum output average value are identified, and the difference between the maximum output average value and the minimum output average value is calculated. Calculated as output uniformity.

  After calculating the output uniformity, the determination unit 36 determines whether the output uniformity is equal to or less than a predetermined determination value. Here, when the output uniformity is equal to or less than the determination value, it is determined to be acceptable. On the other hand, when the output uniformity exceeds the determination value, it is determined as rejected.

  When the determination is finished, the inspection apparatus 30 displays a pass determination and a fail determination on the display unit 27.

  In the step of calculating each block average value, when there is an abnormal block whose output average value is abnormal among a plurality of blocks, the output average values of the abnormal block and the blocks around the abnormal block are compared. When the abnormal block is determined to be a singular point, the abnormal block may be excluded from the inspection target. In this way, abnormal blocks having output values that could not be removed by filtering or block division can be removed from the inspection target, so that the inspection accuracy can be further improved.

FIG. 5 is a diagram showing a plurality of blocks of inspection image data converted into digital data. FIG. 6 is a diagram for explaining the relationship between an abnormal block and blocks around the abnormal block in the inspection image data shown in FIG. FIG. 5 illustrates an example of 10 blocks in the X direction and 10 blocks in the Y direction, that is, 10 × 10 blocks.
The numerical value of each block shown in FIG. 5 is, for example, a sensor output value converted into digital data with 8 bits. Here, the digital value 255 is white and the digital value 0 is black.

When the sensor output is defined so that the digital value becomes 127 at the center of the inspection image data, the abnormal block can be detected as follows.
For example, the block c11 with X = 3 and Y = 2 can obtain only about 1/3 of the output even when compared with the eight surrounding blocks. Such an abnormal block has no continuity with the surrounding blocks, so there is a high possibility that large dust or the like is attached, and this abnormal block can be detected as a singular point.

Here, “continuity” refers to the degree to which the output values of blocks that are continuous in a predetermined direction are present in the predetermined direction while changing the output values with a predetermined slope. As shown in FIG. 6A, when the target block is compared with the blocks around the target block, there is a direction in which the output value changes, and the amount of change in the output value (that is, the slope) is If it is constant, the target block can be said to be continuous with surrounding blocks. The target block in FIG. 6A has an output value of 90. When the change in the X direction is seen, 111 → 90 → 85, and when the change is seen in the Y direction, 113 → 90 → 82. It can be seen that the change gradually decreases.
On the other hand, it can be said that there is no continuity when the output is large for all adjacent blocks, such as the target block of FIG. The target block in FIG. 6B has an output value 146. When the change in the X direction is seen, 107 → 146 (target block) → 121, and when the change is seen in the Y direction, 111 → 146 ( Target block) → 112.

As shown in FIG. 6C, for example, when there is a large difference between the target block and 7 blocks (blocks other than the upper right block in the target block) among the 8 blocks surrounding the target block, Are excluded from the inspection target as singular points.
Alternatively, as shown in FIG. 6 (d), the difference between the target block and 6 blocks (blocks other than the upper right block and right block of the target block) among the 8 blocks around the target block is large, and the target Since the block, the upper right block, and the right block are not continuous, these blocks including the target block are excluded from the inspection target as singular points.

  In the inspection image data shown in FIG. 5, the output value of the block c12 with X = 8 and Y = 6 is about 20% to 30%, which is higher than any surrounding blocks. Since such an abnormal block does not have continuity with any surrounding blocks, there is a high possibility that a correct pixel portion is not formed due to a defect in the manufacturing process, and this abnormal block can be detected as a singular point. it can.

  Furthermore, the output value of the block c13 with X = 2 and Y = 9 is about 20% lower than that of the surrounding blocks, but because of the continuity with the surrounding blocks, this abnormal block is simply used as a singular point. It cannot be detected.

  In this way, an abnormal block can be regarded as a singular point with a high probability when the target block of interest has a significantly different output value compared to all the blocks around the target block. Taking each block of the inspection image data shown in FIG. 5 as an example, if a block having an output difference of about 20% to 30% does not have continuity, the abnormal block can be made a singular point. However, the reference level for determining whether or not an abnormal block is specifically a singular point can be arbitrarily adjusted according to the average change tendency of the in-plane block of each image sensor. Therefore, this reference level is a value deviating from the average optical characteristic of the target image sensor, and a limit value at which an abnormal block is visually recognized can be defined as the reference level. In the above example, an abnormal block having an output difference of about 20% to 30% compared to the surrounding blocks is used as a singular point. However, an output difference of about 50% may be used.

It is a schematic plan view which shows the structure of a solid-state image sensor. It is a block diagram explaining the structure of an inspection apparatus. It is a figure showing typically the state where inspection image data was divided into a plurality of blocks. It is a flowchart which shows the procedure of an inspection method. It is the figure which showed the some block of test | inspection image data as digital data. FIG. 6 is a diagram for explaining a relationship between an abnormal block and blocks around the abnormal block in the inspection image data of FIG. 5.

Explanation of symbols

10 Image sensor 11 Photo diode 12 Vertical CCD (VCCD)
13 Horizontal CCD (HCCD)
30 Inspection Device 33 Smoothing Processing Unit 34 Block Average Value Calculation Unit 35 Uniformity Calculation Unit

Claims (4)

An inspection method of an image sensor for inspecting output uniformity of the image sensor,
Obtaining inspection image data from the image sensor;
Filtering the acquired inspection image data;
Dividing the inspection image data into a plurality of blocks, calculating an average output value of each block,
Calculating output uniformity based on the difference between the maximum output average value and the minimum output average value among the plurality of blocks;
An image sensor inspection method for performing pass / fail determination by comparing the output uniformity with a predetermined determination value. The inspection method according to claim 1,
When there is an abnormal block in which the output average value is abnormal among the plurality of blocks, the output average value of the abnormal block and a block around the abnormal block are compared, and the abnormal block is a singular point. A method for inspecting an image sensor that removes the abnormal block from an inspection target when it is determined that the abnormal block exists. An inspection device for an image sensor that inspects output uniformity of the image sensor,
Inspection image data acquisition means for acquiring inspection image data from the imaging element;
Filtering the acquired inspection image data;
Block average value calculating means for dividing the inspection image data into a plurality of blocks and calculating an average output value of each block;
Uniformity calculating means for calculating output uniformity based on a difference between a maximum output average value and a minimum output average value among the plurality of blocks;
An inspection device for an image sensor, comprising: a determination unit configured to perform a pass / fail determination by comparing the output uniformity with a predetermined determination value. The inspection apparatus according to claim 3,
When there is an abnormal block in which the output average value is abnormal among the plurality of blocks, the output average value of the abnormal block and a block around the abnormal block are compared, and the abnormal block is a singular point. An inspection device for an image sensor that removes the abnormal block from an inspection target when it is determined that the abnormal block exists.

Images