JP2008244637A - Inspection device, solid-state imaging apparatus, and inspection method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To achieve an inspection device for inspecting the quality of a solid-state imaging device by detecting, even stain defects with low contrast at a periphery of an image, which has been impossible to detect by conventional inspection devices. <P>SOLUTION: An inspection circuit 30 calculates a moving average value with a different number of samples in each luminance signal, respectively, from the luminance signals of a horizontal pixel of an image sensor 50 to be inspected. The stain defect is detected, and the quality of the image sensor 50 is determined, depending whether a difference between the moving average values with the different number of samples is not smaller than a prescribed value, and whether the state with the difference continues beyond the predetermined number of states. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an inspection device that inspects the quality of an image sensor such as a CMOS image sensor, an inspection method thereof, and a solid-state imaging device including the inspection device.

  An image sensor as an image pickup apparatus is configured by two-dimensionally collecting a large number of millions or more of pixel portions by applying semiconductor manufacturing technology. This image sensor has undergone many manufacturing processes until completion, and various defects occur due to defects in the manufacturing process.

  One of the defects is “uneven display”. This display unevenness is a change in luminance that occurs with a certain area on the display screen. This display unevenness can be roughly divided into two types: display unevenness that changes gently over the entire screen and display unevenness that changes locally. Hereinafter, the former is referred to as “shading” and the latter as “stain defect”.

  Since this shading gradually changes from the central part to the peripheral part of the pixel, it is often not determined that the display is uneven by the naked eye. On the other hand, since a spot defect is a local luminance change, depending on its size and amount of change, it is determined as a defect that is obvious to anyone's eyes.

  Conventionally, in an inspection apparatus for inspecting the quality of an image sensor, the quality is determined by a process as shown in FIG.

First, output image data (image to be inspected) of the image sensor is taken into an inspection apparatus, and shading correction for removing shading components and smoothing filter processing for removing noise components are performed. After that, edge detection processing for emphasizing spot defect components, block division processing, feature amount calculation processing, and feature amount variation statistical processing were performed, and spot defects were detected to determine whether they were defective (non-patent) Reference 1). A conventional inspection apparatus is also disclosed in Patent Document 1.
JP 2006-135700 A (published May 25, 2006) “Image Sensor Image Quality Inspection Method Using Statistical Error Model” SEMI Technology Symposium 2006 Preliminary Report 4-93

  As described above, in the conventional inspection apparatus, filtering processing such as shading correction and smoothing filter processing is performed before image processing for pass / fail judgment. By this filtering processing, a signal of a spot defect with low contrast in the peripheral portion of the image is obtained. Is corrected, and the signal of the spot defect disappears.

  FIG. 5 shows this state, FIG. 5A shows the luminance signal of the horizontal pixel of the image sensor before the filtering process, and FIG. 5B shows the horizontal pixel after the filtering process. The luminance signal is shown. In the visual inspection, it is determined that there is a spot defect near the horizontal pixel 1400. However, as shown in FIG. 5B, the conventional inspection apparatus generates a spot defect signal near the horizontal pixel 1400 by filtering. Disappeared.

  As described above, in the conventional inspection apparatus, the signal of the spot defect having a low contrast in the peripheral portion of the image is lost by the filtering process. As a result, a defect with low contrast at the periphery of the image cannot be detected, resulting in a problem in that the quality determination accuracy is lowered.

  The present invention has been made in view of the above-described problems, and its object is to detect a solid defect by detecting even a low-contrast spot defect at the periphery of an image, which could not be detected by a conventional inspection apparatus. An object of the present invention is to provide an inspection device that inspects the quality of an element, an inspection method thereof, and a solid-state imaging device including the inspection device.

  In order to solve the above-described problem, the inspection apparatus according to the present invention calculates a moving average value having a different number of samples from an output signal of a solid-state imaging device or a signal obtained by performing predetermined processing on the output signal. The comparison circuit that compares the average calculation circuit with the moving average value of the number of samples calculated by the moving average calculation circuit and outputs a detection signal when the difference between the moving average values is equal to or greater than a predetermined value, and the comparison Whether the solid-state imaging device is good or bad is based on whether the counter circuit that counts the detection signal of the circuit and the output signal of the counter circuit are continuous and the number of output signals is a predetermined number or more. And a determination circuit for determining whether or not.

  According to the above configuration, the inspection apparatus according to the present invention uses the fact that a spot defect has a certain area or more and has a contrast difference in pixel value as compared with the pixels in the peripheral region. Each of the moving average values different from each other, there is a difference of a predetermined value or more between the moving average values of the different number of samples, and whether there is a predetermined number or more of the state where there is a difference, A spot defect in the solid-state imaging device to be inspected is detected and the quality is determined.

  With the above inspection apparatus, it is not necessary to perform filtering processing such as shading correction and smoothing filter processing that has been performed in conventional inspection apparatuses, so in addition to high-contrast spot defects, low contrast in the periphery of the image Spot defects can also be detected. As a result, there is an effect that it is possible to realize an inspection apparatus that can detect a defect of low contrast in the peripheral portion of the image and inspect the quality of the solid-state imaging device, which cannot be detected by the conventional inspection apparatus. .

  In addition, since the conventional inspection apparatus detects a spot defect and determines whether it is good or bad by image processing, a large-capacity memory for image processing is required, which makes the inspection apparatus large-scale. It was. On the other hand, the inspection apparatus according to the present invention detects spot defects and determines pass / fail judgment as described above, and therefore has the effect of reducing the size of the inspection apparatus without requiring a large-capacity memory.

  In the inspection apparatus according to the present invention, the number of samples of the moving average calculation circuit is preferably a power of two.

  If the number of samples in the moving average calculation circuit is a power of 2, the division at the time of moving average calculation in the moving average calculation circuit only requires a bit shift, so that the circuit can be further reduced in size.

  The inspection apparatus according to the present invention preferably further includes a storage circuit for storing the sample value of the moving average calculation circuit, the predetermined level value of the comparison circuit, and the predetermined value of the determination circuit.

  According to said structure, since each said numerical value can be changed, there exists the further effect that a suitable test | inspection can be performed suitably, for example, the test | inspection corresponding to the individual difference of a solid-state image sensor, for example. Note that the sample value information stored in the storage circuit is also passed to the solid-state imaging device and used for setting a reference pixel, which will be described later.

  In the inspection apparatus according to the present invention, when a pixel that is more than the number of samples of the moving average calculation circuit from the pixel at the end of the horizontal pixel of the solid-state image sensor is used as a reference pixel, the reference pixel is shifted to the pixels at both ends. When the solid-state imaging device is configured to output the output signals in order, it is preferable that the inspection apparatus can be handled by one apparatus.

  When the moving average value is calculated by the moving average calculation circuit, from the pixel at the end of the horizontal pixel of the solid-state imaging device to the pixel of the moving average calculation circuit minus the first sample (hereinafter referred to as the first pixel). The moving average cannot be calculated. Accordingly, there is a possibility that the spot defect of the top pixel cannot be detected as it is. Therefore, when a pixel that is more than the number of samples of the moving average calculation circuit from the pixel at the end of the horizontal pixel of the solid-state image sensor is used as the reference pixel, the output signal is output in order from the reference pixel to the pixels at both ends. Thus, a solid-state imaging device is configured. As a result, the moving average value can be calculated even at the top pixel, and a spot defect can be detected.

  Here, even in such a case, that is, when output signals are output in two systems from the solid-state imaging device, the inspection apparatus according to the present invention can be handled by one apparatus. Thereby, compared with the case where two inspection apparatuses are provided, an inspection cost can be reduced. Further, when the solid-state imaging device and the inspection apparatus are contained in one chip, there is an additional effect that the chip size can be reduced.

  In order to solve the above problems, a solid-state imaging device according to the present invention includes a solid-state imaging element and the inspection device.

  As a result, the solid-state imaging device according to the present invention is an inspection device that inspects the quality of the solid-state imaging device by detecting a spot defect having a low contrast in the peripheral portion of the image, which could not be detected by a conventional inspection device. The solid-state imaging device provided can be provided.

  In addition, for example, after performing a spot defect inspection by a conventional inspection method (transferring image data of a solid-state imaging device to an external inspection apparatus, performing image calculation processing, and determining), the condition of the light source is further changed. When performing a spot defect inspection with low contrast at the periphery of the image, the solid-state imaging device includes the inspection device, so that processing for transferring data to the outside is unnecessary as compared with the conventional inspection method. Therefore, there is a further effect that the inspection time can be shortened, and thus the inspection cost can be reduced.

  An inspection method according to the present invention is an inspection method for inspecting the quality of a solid-state image sensor to solve the above-described problem, and is a signal obtained by performing a predetermined process on an output signal of the solid-state image sensor or on the output signal. From the calculation step of calculating the moving average value with different number of samples, the moving average value with different number of samples calculated in the calculation step is compared, and detected when the difference of moving average value is more than a predetermined level Whether the comparison step for outputting a signal, the counting step for counting the detection signal output in the comparison step, and the counting in the counting step are continuous, and the count number is a predetermined number or more. A determination step of determining whether the solid-state imaging device is good or not based on whether or not the solid-state imaging device is good.

  According to the above configuration, the inspection method according to the present invention uses the fact that the spot defect has a certain area or more and has a contrast difference in pixel value as compared with the pixels in the peripheral region. Each of the moving average values different from each other, there is a difference of a predetermined value or more between the moving average values of the different number of samples, and whether there is a predetermined number or more of the state where there is a difference, A spot defect in the solid-state imaging device to be inspected is detected and the quality is determined.

  With the above inspection method, it is not necessary to perform filtering processing such as shading correction and smoothing filter processing that has been performed in conventional inspection apparatuses, so in addition to high-contrast spot defects, low contrast in the periphery of the image Spot defects can also be detected. Accordingly, there is an effect that it is possible to realize an inspection method for inspecting the quality of the solid-state imaging device by detecting a spot defect having a low contrast in the peripheral portion of the image, which could not be detected by a conventional inspection apparatus. .

  An inspection apparatus according to one of the present invention includes a moving average calculation circuit that calculates a moving average value having a different number of samples from an output signal of a solid-state imaging device or a signal obtained by performing predetermined processing on the output signal, A comparison circuit that compares moving average values of different numbers of samples calculated by the moving average calculation circuit and outputs a detection signal when the difference in moving average values is equal to or greater than a predetermined value, and a detection signal of the comparison circuit A counter circuit that counts and a determination circuit that determines whether the solid-state imaging device is good or not based on whether the output signals of the counter circuit are continuous and the number of output signals is a predetermined number or more. It is characterized by having.

  As a result, there is an effect that it is possible to realize an inspection apparatus that can detect a defect of low contrast in the peripheral portion of the image and inspect the quality of the solid-state imaging device, which cannot be detected by the conventional inspection apparatus. .

  The embodiment of the present invention is described below with reference to FIGS.

  FIG. 1 shows an example of a configuration of a CMOS image sensor (solid-state imaging device, solid-state imaging device) (hereinafter simply referred to as an image sensor) 50.

  As illustrated, the image sensor 50 includes a pixel unit 12 in which a plurality of pixels 11 are arranged in a two-dimensional matrix, a vertical scanning circuit 13, a horizontal scanning circuit 14, a column signal circuit 15, a horizontal signal line 16, and an output circuit. 17, a switch 18, a selection transistor Tr, and an inspection circuit (inspection device) 30. The operation will be briefly described. First, when light enters the pixel 11, a photoelectrically converted charge is generated in accordance with the intensity of the photodiode (not shown) of the pixel 11, and the charge is converted into a voltage signal by an amplifier circuit (not shown) of the pixel 11. Converted. Next, when the pixel 11 is selected by the vertical scanning circuit 13, the voltage signal is supplied to the column signal circuit 15, and the column signal circuit 15 performs noise cancellation processing. Thereafter, when the selection transistor Tr is turned on by the horizontal scanning circuit 14, the voltage signal output from the column signal circuit 15 is applied to the output circuit 17 via the selection transistor Tr and the horizontal signal line 16, and processing such as amplification is performed. Output. The switch 18 connects the output circuit 17 and the inspection circuit 30 at the time of inspection based on a control signal given from the outside.

  FIG. 2 shows the configuration of the inspection circuit 30. The inspection circuit 30 is a circuit that inspects the quality of the image sensor 50 based on whether or not a spot defect has occurred. Note that a spot defect is a difference in contrast between the pixel values of pixels having a certain area or more in the image formed by a signal output from the image sensor as compared with the pixels in the peripheral region.

  The inspection circuit 30 includes an L point moving average calculation circuit 22, an M point moving average calculation circuit 23, a comparison circuit 24, a counter circuit 25, and a determination circuit 26.

  The L point moving average calculation circuit 22 calculates the L point (number of samples) moving average value of each luminance signal from the luminance signal (output signal) of the pixel 11 of the image sensor 50. The M point moving average calculation circuit 23 calculates an M point (number of samples) moving average value of each luminance signal from the luminance signal.

For example, when the above moving average values are 8-point moving average values, the coordinates of the pixel 11 are set to N-7, N-6, N-5, N-4, N-3, N-2, N-1, N The luminance signal of each pixel is S (N-7), S (N-6), S (N-5), S (N-4), S (N-3), S (N-2 ), S (N-1), S (N)
8-point moving average of coordinate N = {S (N-7) + S (N-6) + S (N-5) + S (N-4) + S (N-3) + S (N-2) + S (N- 1) + S (N)} / 8
Calculate as follows. Needless to say, this calculation method is an example.

  3 shows an L point (here, L = 8) moving average value (thin line) calculated by the L point moving average calculating circuit 22 and an M point (here, M) calculated by the M point moving average calculating circuit 23. = 64) Moving average value (thick line). The horizontal axis of the graph in the figure is the pixel 11 of the image sensor 50, the vertical axis is the luminance of each pixel, and the luminance signal of each pixel is indicated by using rhombus marks.

  The comparison circuit 24 compares the L point moving average value calculated by the L point moving average calculation circuit 22 with the M point moving average value calculated by the M point moving average calculation circuit 23. Specifically, for example, the L-point moving average value and the M-point moving average value of the luminance signal of the pixel given “a” shown in FIG. 1 are compared. In the comparison, a detection signal is output when the difference between the moving average values is a predetermined value or more (for example, 50 or more).

  The counter circuit 25 counts the detection signal output from the comparison circuit 24. The determination circuit 26 determines the quality of the image sensor 50 based on whether the output signals of the counter circuit 25 are continuous and the number of output signals is a predetermined number or more (for example, 5 or more). To do. Specifically, if the output signals of the counter circuit 25 are continuous and the number of output signals is equal to or greater than a predetermined number, it is determined that there is a spot defect and the image sensor 50 is defective.

  By the way, in the image sensor 50, it is determined that there is a spot defect in the vicinity of the horizontal pixel 1400 in the visual inspection. Here, in FIG. 2 showing the L point moving average value calculated by the L point moving average calculation circuit 22 and the M point moving average value calculated by the M point moving average calculation circuit 23, in the vicinity of the horizontal pixel 1400. Thus, it can be seen that there is a large difference between the L point moving average value and the M point moving average value, and the large difference continues to the vicinity of the horizontal pixel 1450. Therefore, the inspection circuit 30 can reliably detect a spot defect near the horizontal pixel 1400 that has been found in the visual inspection.

  The inspection circuit 30 is configured as described above, calculates moving average values with different numbers of samples, and there is a difference of a predetermined value or more between the moving average values with different numbers of samples. Depending on whether or not there are a predetermined number or more of continuous states, the detection of spot defects and the quality of the image sensor to be inspected are determined. With this configuration, it is not necessary to perform filtering processing such as shading correction and smoothing filter processing that has been performed in conventional inspection apparatuses, so in addition to high-contrast stain defects, low-contrast stains at the periphery of the image. Defects can also be detected.

  In addition, since the conventional inspection apparatus detects a spot defect and determines whether it is good or bad by image processing, a large-capacity memory for image processing is required, which makes the inspection apparatus large-scale. It was. On the other hand, the inspection circuit 30 can detect a spot defect and determine whether it is good or bad by processing such as successive comparison of moving average values, so that the circuit can be downsized.

  The number of samples in each moving average calculation circuit 22, 23 is arbitrary but is preferably a power of two. If it is a power of 2, the division at the time of calculating the moving average need only be a bit shift, so that the circuit can be reduced in size.

  Further, the inspection circuit 30 may include a memory for storing the sample values of the moving average calculation circuits 22 and 23, the predetermined value of the comparison circuit 24, and the predetermined value of the determination circuit 26, or the above-described circuits. May be provided respectively. In this case, since each of the above numerical values can be changed, for example, an appropriate inspection such as an inspection corresponding to individual differences of image sensors can be performed.

  Here, the luminance signal is used. However, the present invention is not limited to this. For example, a color difference signal (a signal subjected to a predetermined process) may be used.

  By the way, when moving average values are calculated by the moving average calculation circuits 22 and 23, the pixels of the sample number minus −1 of the moving average calculation circuits 22 and 23 from the pixel at the end of the pixel 11 of the image sensor 50 (hereinafter referred to as the pixels) The moving average value cannot be calculated until the first pixel. Accordingly, there is a possibility that the spot defect of the top pixel cannot be detected as it is. Therefore, when a pixel that is more than the number of samples of the M-point moving average calculation circuit 23 from the pixel at the end of the pixel 11 of the image sensor 50 is set as a reference pixel, a luminance signal (horizontal signal) is sequentially transmitted from the reference pixel to the pixels at both ends. ) Is configured so that the horizontal scanning circuit 14 of the image sensor 50 is configured. FIG. 4 shows the concept of an image sensor that outputs output signals in these two systems.

  The configuration in which the output signals are output by the two systems will be specifically described using the pixel with “a” shown in FIG. 1 as the reference pixel. When the pixels a to f in the figure are selected by the vertical scanning circuit 13, the selection transistor Tr to which “a” corresponding to the pixel a is attached is turned on by a signal from the horizontal scanning circuit 14, and the horizontal signal line 16 is turned on. The luminance signal of the pixel a is output to Next, by sequentially selecting the selection transistor Trb to the selection transistor Trf, the luminance signals of the pixels b to f are sequentially output to the horizontal signal line 16, and the configuration in which the output signals are output in the above two systems is achieved. With this configuration, it is possible to calculate the moving average value even at the top pixel and to detect a spot defect. The reference pixel is set by the sample value of the M-point moving average calculation circuit 23 stored in the memory.

  Here, even in the case of a configuration in which output signals are output by the two systems described above, the inspection circuit 30 can be handled by a single circuit. Thereby, the chip size of the image sensor 50 can be reduced as compared with the case where two inspection circuits are provided.

  As the spot defect inspection of the image sensor 50, after performing inspection by a conventional inspection method (transferring image data of the image sensor to an external inspection apparatus and performing image arithmetic processing to determine pass / fail), further inspection is performed. It is preferable to use a technique for inspecting a stain defect having a low contrast in the peripheral portion of the image by the circuit 30. At the time of the re-inspection, since the image sensor 50 includes the inspection circuit 30, it is not necessary to perform processing such as transferring data to the outside as compared with the conventional inspection method. Inspection costs can be reduced.

  The present invention is not limited to the above-described embodiments, and various modifications can be made within the scope shown in the claims, and embodiments obtained by appropriately combining technical means disclosed in the embodiments are also included. It is included in the technical scope of the present invention.

  The inspection apparatus according to the present invention is suitable as an inspection apparatus for inspecting the quality of an image sensor such as a CMOS image sensor.

It is a figure which shows the structure of the CMOS image sensor which concerns on one Embodiment of this invention. It is a block diagram which shows the structure of the test | inspection circuit which concerns on one Embodiment of this invention. It is a graph which shows the moving average value calculated in the moving average calculation circuit. It is a conceptual diagram of the image sensor which outputs an output signal by two systems. It is a figure which shows the processing flow of the conventional inspection apparatus. It is a graph which shows the luminance signal of the horizontal direction pixel of an image sensor, Drawing 6 (a) shows before filtering processing, and Drawing 6 (b) shows after filtering processing.

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 Pixel 12 Pixel part 13 Vertical scanning circuit 14 Horizontal scanning circuit 15 Column signal circuit 16 Horizontal signal line 17 Output circuit 18 Switch 22 L point moving average calculation circuit (moving average calculation circuit)
23 M-point moving average calculation circuit (moving average calculation circuit)
24 comparison circuit 25 counter circuit 26 determination circuit 30 inspection circuit (inspection apparatus)
50 CMOS image sensor (solid-state imaging device, solid-state imaging device)

Claims (6)

A moving average calculation circuit for calculating a moving average value having a different number of samples from an output signal of the solid-state imaging device or a signal obtained by performing predetermined processing on the output signal;
A comparison circuit that compares the moving average values of different numbers of samples calculated by the moving average calculation circuit, and outputs a detection signal when the difference of the moving average values is equal to or greater than a predetermined value;
A counter circuit for counting detection signals of the comparison circuit;
A determination circuit that determines whether the solid-state imaging device is good or not based on whether the output signals of the counter circuit are continuous and whether the number of output signals is a predetermined number or more. Inspection device characterized by   2. The inspection apparatus according to claim 1, wherein the number of samples of the moving average calculation circuit is a power of two.   3. The inspection apparatus according to claim 1, further comprising a storage circuit that stores a sample value of the moving average calculation circuit, a predetermined value of the comparison circuit, and a predetermined value of the determination circuit. Inspection equipment. When a pixel that is more than the number of samples of the moving average calculation circuit from the pixel at the end of the horizontal pixel of the solid-state image sensor is used as a reference pixel, the output signal is output in order from the reference pixel to the end pixels on both sides. When the above solid-state imaging device is configured,
The inspection apparatus according to claim 1, wherein the inspection apparatus can be handled by one apparatus. A solid-state image sensor;
A solid-state imaging device comprising: the inspection device according to claim 1. An inspection method for inspecting the quality of a solid-state image sensor,
A calculation step for calculating a moving average value having a different number of samples from an output signal of the solid-state imaging device or a signal obtained by performing a predetermined process on the output signal;
A comparison step of comparing the moving average values of the number of samples calculated in the calculation step, and outputting a detection signal when the difference of the moving average values is equal to or greater than a predetermined level;
A counting step for counting the detection signals output in the comparison step;
And a determination step of determining whether the solid-state imaging device is good or not based on whether the count in the counting step is continuous and whether the count number is a predetermined number or more. Inspection method.

Images