JP2009276108A - Image sensor surface inspection method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform uniform inspection on a CCD sensor surface which could not be implemented with conventional constitutions. <P>SOLUTION: Silicon fragments, formed by dicing and an organic substance, are used as objects for foreign matters adhering onto the image sensor (CCD sensor) surface, and inspection for detecting them is performed. As an inspection method, a method is used, where a master sample is prepared for each type of inspection objects, and inspection is performed, by determining the difference in the luminance distribution of the master sample. An inspection device 13 used therefor is constituted of a light source part 14, an auxiliary light source part 15, a stage part 16, an optical system 17, and an image processing system 18. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an inspection of a surface of an image sensor.

  A production process of a conventional CCD sensor (image sensor) will be described with reference to FIGS.

  FIG. 7A is a schematic diagram showing a first process of production of a CCD sensor, FIG. 7B is a schematic diagram showing a second process of production of a CCD sensor, and FIG. 1 is a schematic view showing a CCD sensor.

  FIGS. 7A to 7C briefly describe the production process of the CCD sensor.

  First, the CCD sensor 1 is obtained by individually dividing a semiconductor wafer 3 attached on a dicing sheet 2 by dicing. The divided CCD sensors 1 are individually separated and stored in the tray 4 (FIG. 7A).

  Subsequently, the CCD sensor 1 is taken out from the tray 4, and bumps for electrical connection are provided by the bump bonder 5, and then stored in the tray 4 again (FIG. 7B).

  Subsequently, the bumped CCD sensor 1 housed in the tray 4 is connected to the substrate 6 via a glass assembly 7 to complete the CCD camera 8 (FIG. 7C).

  In the CCD sensor production line process described above, the semiconductor wafer 3, the CCD sensor 1 after the semiconductor wafer 3 is diced, the CCD sensor 1 after being placed on the tray 4, and the CCD sensor 1 being rearranged on the tray 4 after bump formation. , It is necessary to perform surface inspection for each. This is because, in each of the above-described processes, foreign matter such as silicon debris generated during dicing or organic matter entering from the outside may adhere to the semiconductor wafer 3 or the CCD sensor 1 and the performance as the CCD camera 8 may deteriorate. is there.

  In particular, in the case of a CCD sensor, the recent trend toward miniaturization, such as the pixel size being changed from 10 μm to 2 μm, may cause trouble in image reading even if a minute foreign matter of the order of 1 μm adheres. Since it is a defective product, it is necessary to inspect every operation.

  As this foreign matter inspection device on the surface of the CCD sensor, there is one that includes an imaging camera and automatically inspects the appearance by image recognition (see, for example, Patent Document 1).

  FIG. 8 is a schematic diagram of a conventional surface inspection apparatus described in Patent Document 1. In FIG.

In FIG. 8, the inspection apparatus 9 includes an imaging camera 10, and the CCD sensor 12 whose position is adjusted by the XY stage 11 is imaged by the imaging camera 10, and the captured image is recognized, whereby the CCD automatically The appearance inspection (surface foreign matter inspection) of the sensor 12 is performed.
JP 2007-003326 A

  However, the image sensor (CCD sensor) may have a non-uniform reflection state on the entire surface. In this case, with respect to white light in which RGB components are mixed at an approximately equal ratio, the RGB components have different reflectances, and the contrast and hue of the reflected light vary greatly depending on the position of the image sensor surface. For this reason, the conventional configuration described above has a problem that accurate inspection cannot be performed on the surface of such an image sensor.

  The present invention is for solving the above-described conventional problems, and an object thereof is to enable an accurate inspection regardless of the reflection condition of the image sensor surface.

  In order to achieve the above object, an image sensor surface inspection method according to the present invention includes a first step of dividing a captured image obtained by capturing the surface of an image sensor with an imaging unit into a plurality of regions, and each wavelength in each of the plurality of regions. A second step of measuring the luminance of the first, a third step of detecting light having a wavelength at which the luminance is outside the dynamic range of the imaging means, a fourth step of correcting the luminance of the light of the detected wavelength, And a fifth step of inspecting the surface of the image sensor based on the corrected captured image.

  As described above, according to the present invention, an accurate inspection that does not depend on the reflection condition of the image sensor surface can be performed.

  Embodiments of the present invention will be described with reference to the drawings. In the following description, the same components are denoted by the same reference numerals and description thereof is omitted.

(Embodiment 1)
FIG. 1 is a schematic configuration diagram of an inspection apparatus according to Embodiment 1 of the present invention.

  In the present embodiment, silicon fragments and organic substances generated by dicing as foreign matters attached on the surface of an image sensor (CCD sensor) are targeted, and an inspection for detecting them is performed.

  In addition, as an inspection method, a method is used in which a master sample is prepared for each type of inspection target and an inspection is performed by obtaining a difference from the luminance distribution of the master sample.

  In FIG. 1, the inspection apparatus 13 includes a light source unit 14, an auxiliary light source unit 15, a stage unit 16, an optical system 17, and an image processing system 18. In the present embodiment, the light from the light source unit 14 is irradiated onto the object to be measured on the stage unit 16 via the optical system 17, reflected by the object to be measured, and again the light that has passed through the optical system 17. Imaging is performed by the image processing system 18 (imaging means). At this time, the measurement object is irradiated with light from the auxiliary light source 15 as auxiliary light as necessary.

  Subsequently, these components will be described in detail.

  The light source unit 14 includes a white lamp excitation light source 19, and a monochrome (monochrome) gradation filter 20, an R (red) gradation filter 21, and a G (green) gradation as means for adjusting light from the white lamp excitation light source 19. The filter 22 is composed of a B (blue) gradation filter 23 and filter moving mechanisms 24 to 27 for independently moving the gradation filters 20 to 23.

  The auxiliary light source unit 15 includes an R (red) diode light source 28, a G (green) diode light source 29, and a B (blue) diode light source 30.

  The stage unit 16 includes a stage 31 and a scanning motor 32. On the stage 31, a tray 34 in which a CCD sensor 33 (image sensor) that is an object to be inspected is placed. The stage 31 moves in the line scan sub-scanning direction (horizontal direction on the paper surface) indicated by A in FIG.

  The optical system 17 includes a half mirror 35 for guiding light from the light source unit 14 to the stage unit 16 and a half mirror 36 for guiding light from the auxiliary light source unit 15 to the stage unit 16.

  The image processing system 18 includes a microscope (not shown), a line scan camera 37, and an image processing unit 38 to which an image captured by the line scan camera 37 is transferred and processed. The line scan camera 37 scans in a line shape with respect to the main line scanning direction (direction perpendicular to the paper surface) indicated by B in FIG. The line scan camera 37 controls the exposure time of the imaging CCD element (described later in FIG. 6) of the line scan camera 37 with respect to the illumination light combined with the light from the light source unit 14 and the auxiliary light source unit 15. An exposure time adjustment unit (not shown) for controlling the gain and a gain adjustment unit (not shown) for controlling the gain.

  The inspection device 13 includes a controller 39 (control device) having an electric control unit for controlling the diode light sources 28 to 30 therein. The controller 39 includes the filter moving mechanisms 24 to 27 and the diode light source 28. ˜30, the scanning motor 32, and the line scan camera 37 are connected to each other and can be operated in synchronization with each other.

  Here, the gradation level of the gradation filters 20 to 23 for monochrome and each color will be described using the gradation filter 20.

  FIG. 2 is a diagram showing the density of the gradation filter 20.

  In FIG. 2, the gradation filter 20 has transmission regions 20a to 20f having uniform transmittance in each region. The transmissive regions 20a to 20f have a structure in which the light transmittance gradually increases from the transmissive region 20a to the transmissive region 20f. In the present embodiment, the width of each of the transmission regions 20a to 20f is at least larger than the width of the light beam of the white lamp excitation light source 19. This is for making the light from the white lamp excitation light source 19 through the transmission regions 20a to 20f as uniform as possible.

  In FIG. 2, the gradation filter 20 has been described on behalf of the gradation filters 20 to 23, but the gradation filters 20 to 23 have the same structure except for the difference between monochrome and RGB colors. It has become.

  Next, the light adjustment parameter determination method and the CCD sensor surface inspection method in this embodiment will be described with reference to flowcharts.

  FIG. 3 is a diagram illustrating a flowchart for determining light adjustment parameters for inspection according to the first embodiment, and FIG. 4 is a diagram illustrating a flowchart for surface inspection according to the first embodiment.

  In FIG. 3, first, the required number of line scan imaging (number of scan lines) is calculated from the dimensions of the CCD sensor 33 to be inspected and the alignment mark position. Then, the start position of each line scan imaging is determined, and is stored in the controller 39 as a light adjustment parameter for each scan line (step S1).

  Next, the luminance distribution is grasped for each scan line determined in step S1 (step S2).

  Next, based on the state of the luminance distribution in step S2, a luminance adjustment position that requires luminance adjustment because the luminance is different from the others is determined for each scan line (step S3).

  Next, the field of view taken by the line scan camera 37 is moved to the brightness adjustment position determined in step S3 (step S4).

  Next, in the imaging field of view in step S4, the gradation filters 20 to 23 are moved by the filter moving mechanisms 24 to 27 to change the transmittance and adjust the luminance. For example, when the R (red) component is larger than the other two components in step S2, the brightness adjustment is performed by gradually decreasing the R component by moving the gradation filter 21 in steps for each region. Thus, adjustment is made so that the luminance values of the three RGB colors are closest to each other (step S5).

  Next, it is checked whether or not the luminance after the adjustment in step S5 is within the luminance range that can be inspected by the inspection device 13 (step S6).

  Next, when it becomes in the brightness | luminance range which can be test | inspected by step S6, the brightness | luminance adjustment position and adjustment parameter (movement amount of the filter moving mechanisms 24-27) are stored in the memory | storage part (not shown) of the inspection apparatus 13. FIG. Store (step S7).

  If the luminance is not within the inspectable luminance range in step S6, the luminance is improved using the diode light sources 28 to 30 if the luminance is smaller than the inspectable luminance, and the luminance is increased using the gradation filter 20 if the luminance is larger than the inspectable luminance. Decrease (step S8).

  Next, the brightness adjustment position and the adjustment parameters (the movement amounts of the filter moving mechanisms 24 to 27 and the set values of the electric control unit) in step S8 are stored in the storage unit of the inspection apparatus 13 (step S9).

  The operations in steps S4 to S9 described above are performed for all the brightness adjustment positions on the surface of the CCD sensor 33 (step S10).

  Here, in order to determine the reference for brightness adjustment, the reflectance of the CCD sensor 33 in the present embodiment under various factors will be examined.

  First, Rc, Rs, and Ro were the light reflectivities when white light was irradiated to the surface of the CCD sensor 33, silicon fragments generated in the pre-inspection process, and organic matter, respectively. And it discovered that those reflectances established the relationship of following formula (1) in this Embodiment.

  Using the relationship of the above formula (1), in the present embodiment, when silicon debris is present on the surface of the CCD sensor 33 as the object to be inspected, it is detected as white foreign matter, and when organic matter is present, it is detected as black foreign matter. To do. Therefore, the upper and lower limit threshold values for the luminance that the surface of the CCD sensor 33 is considered to be imaged (no foreign matter is present) are determined in advance for the imaging result of the line scan camera 37. Furthermore, a lower limit threshold value of an area that is regarded as having a foreign object on the surface of the CCD sensor 33 is determined in advance based on the pixel size of the CCD sensor 33 and the like.

  Next, specific luminance adjustment criteria will be described. As a reference for brightness adjustment, in order to detect both white foreign matter and black foreign matter, the brightness of the surface of the CCD sensor 33 in the absence of foreign matter is set to the median value of the dynamic range of the line scan camera 37 (8-bit 255 gradation). If there is, it is desirable to adjust to around 128). In this case, the upper threshold is set to 3/2 of the median of the dynamic range of the line scan camera 37, and the lower threshold is set to ½ of the median of the dynamic range of the line scan camera 37.

  Here, for simplicity of explanation, the luminance threshold is described as an absolute value with respect to the luminance. However, the luminance of the surface of the CCD sensor 33 in a state where there is no foreign object is obtained, and the relative value may be used as the threshold. good. In addition, the simplest technique is to binarize and label the image, and to consider a single label area as a single foreign object, but it is possible to combine labels of a certain distance or less to process a single label. A label having a certain area or less may be erased as noise.

  In FIG. 4, first, the end of line scan imaging is confirmed (step S11).

  Next, image processing such as binarization labeling is performed on the captured image (step S12).

  Next, the presence / absence of an area that is equal to or greater than the upper limit threshold of luminance is confirmed in the imaging area (step S13).

  Next, if there is a region equal to or greater than the upper limit threshold of luminance in step S13, it is confirmed whether the area of the region is equal to or greater than the lower limit threshold of area (step S14).

  Next, when the area is equal to or larger than the lower limit threshold of the area in step S14, the area is determined to be a white foreign object (step S15).

  Further, when there is no region equal to or higher than the upper limit threshold of luminance in step S13 and when it is smaller than the lower limit threshold of area in step S14, the presence / absence of an area equal to or lower than the lower limit threshold of luminance is confirmed (step S16).

  Next, if there is a region below the lower limit threshold of luminance in step S16, it is checked whether the area of the region is equal to or greater than the lower limit threshold of area (step S17).

  Next, when the area is equal to or larger than the lower limit threshold of the area in step S17, the area is determined to be a black foreign object (step S18).

  Next, when there is no region below the lower limit threshold of luminance in step S16 and when the area is smaller than the lower limit threshold of area in step S17, it is determined that there is no foreign matter in which neither white foreign matter nor black foreign matter exists (step S19).

  And after performing step S11-S19, it moves to the following line scan image, and performs the same process (step S20).

  In the present embodiment, the determination of the white foreign matter (steps S13 to S15) and the determination of the black foreign matter (steps S16 to S18) are the same, but the order of these determinations is reversed. The effect of can be produced.

  An imaging result obtained by imaging the surface of the CCD sensor 33 according to the above-described flowchart will be described.

  FIG. 5 is a schematic diagram of an imaging result in the inspection according to the first embodiment. Specifically, in the present embodiment, the imaging result obtained by imaging the surface of the CCD sensor 33 by the line scan camera 37 under illumination conditions where the white lamp excitation light source 19 has a constant light amount is shown. Here, the foreign substances 40 and 41 are both white foreign substances.

  The imaging results in this embodiment are shown in Table 1 below for the average luminance in the region for each luminance distribution of RGB for each region shown in FIG. The numbers in the table indicate gradations in the case of 8 bits (for example, R is 255 gradations in the region 33c).

  From FIG. 5 and Table 1, it can be seen that in the regions 33b, 33d, 33e, 33f, and 33h on the surface of the CCD sensor 33, an image capable of detecting white foreign matter is obtained. On the other hand, in the regions 33a, 33c, 33g, and 33i, the luminance is saturated as shown in FIG. 5 and Table 1, and it can be seen that it is difficult to detect the foreign matter 40 (white foreign matter).

  The inventors have considered the areas 33a, 33c, 33g, and 33i by analyzing the luminance distribution. In these areas, only the R component of RGB exceeds the dynamic range of the line scan camera 37. I found out.

  Based on this knowledge, in the present embodiment, only the R component of the white light from the white lamp excitation light source 19 is partially blocked at the timing of imaging these regions 33a, 33c, 33g, and 33i, and line scanning is performed. The dynamic range of the camera 37 could be suppressed.

  Next, a method for realizing this will be described.

  In the actual inspection of the CCD sensor 33, in consideration of the inspection tact time, the main scan per line of the line scan requires about 0.1 seconds. However, in the general white lamp excitation light source 19, it is very difficult to change the color tone only by the electric control means at this time. For this reason, in the present embodiment, as shown in FIG. 1, the color tone and the amount of light of the illumination light are dynamically changed using the gradation filters 20 to 23 during the main scan of the line scan imaging.

  Here, supplementary explanation will be given for the main scanning direction and the sub-operation direction in the present embodiment.

  FIG. 6 is an enlarged view of a main part of the stage 31 and the line scan camera 37. This is an enlarged view of the main parts of the stage 31 and the line scan camera 37 of the present embodiment, and is a view of these main parts viewed from the same direction as FIG.

  In FIG. 6, the stage 31 moves in the direction (sub-scanning direction) indicated by A in FIG. 1, and the imaging CCD element 42 in the line scan camera 37 is CCD-transferred in the opposite direction to A.

  At the time of inspection in an actual production line, the light adjustment parameter for each start position of the line scan set by the above method is changed in synchronization with the position of the surface of the CCD sensor 33 during the main scan of the line scan. In addition, the contrast and hue changes due to the position of the surface of the CCD sensor 33 are effectively suppressed.

  Further, in order to alleviate uneven illumination by the gradation filters 20 to 23, a line subjected to Time Delay Integration processing, which is a technique for synchronizing the charge transfer timings of the stage 31 and the imaging CCD element 42 and integrating the captured images. It is desirable to perform scan imaging.

  In the present embodiment, the diode light sources 28 to 30 are used as the auxiliary light source unit 15. However, when such a diode light source is used, the illumination light combined with the white lamp excitation light source 19 is the wavelength of the diode light source. The spectrum light has a peak. Therefore, it is desirable not to use the auxiliary light source unit 15 as much as possible when the luminance adjustment is possible only with the gradation filters 20 to 23.

  In the present embodiment, a halogen lamp is used as a specific example of the white lamp excitation light source 19, but it is also conceivable to use other lamps. However, when a metal halide lamp or mercury lamp is used, the luminance is high, but it is disadvantageous in terms of running cost, etc. When using a white LED, it is advantageous in running cost but advantageous in terms of luminance. hard. Actually, it is only necessary to make decisions based on the wavelength components and the background pattern, based on these principles. Here, a microscope is used as an example of an optical system, but an optical system other than a microscope can be configured, and the situation may be different in that case.

  In this embodiment, since the color filter of the CCD sensor 33 that is the object to be inspected is a CMY complementary color system, the RGB primary color system and the monochrome gradation filter are used. However, the color filter of the CCD sensor 33 is RGB. In the case of the primary color system, a CMY complementary color system and a monochrome gradation filter are used. That is, the supplement of the color filter of the CCD sensor 33 is used.

  When the luminance is lowered by the white lamp excitation light source 19 (for example, when the luminance value is 64 which is half of 128), the S / N (Signal / Noise) is doubled. By using only the diode light sources 28 to 30 without using the light source 19, more accurate inspection can be performed.

  According to the present invention, uniform foreign matter inspection can be performed regardless of the surface position of the object to be inspected, which is useful for, for example, an inspection apparatus for inspecting foreign matter attached to the surface of a CCD sensor.

Schematic configuration diagram of the inspection apparatus of the first embodiment The figure which shows the shading of the gradation filter 20 Flowchart for determining light adjustment parameters for inspection according to Embodiment 1 Flow chart of surface inspection according to Embodiment 1 Schematic diagram of imaging results in inspection of embodiment 1 Enlarged view of main parts of stage 31 and line scan camera 37 (A) Schematic diagram showing first step of production of CCD camera, (b) Schematic diagram showing second step of production of CCD camera, (c) Schematic diagram showing CCD camera Schematic of the conventional surface inspection apparatus described in Patent Document 1

Explanation of symbols

DESCRIPTION OF SYMBOLS 13 Inspection apparatus 14 Light source part 15 Auxiliary light source part 16 Stage part 17 Optical system 18 Image processing system 19 White lamp excitation light source 20-23 Gradation filter 24-27 Filter moving mechanism 28-30 Diode light source 31 Stage 32 Scanning motor 33 CCD sensor 34 Tray 35, 36 Half mirror 37 Line scan camera 38 Image processing unit 39 Controller

Claims (5)

A first step of dividing a captured image obtained by imaging the surface of the image sensor with an imaging unit into a plurality of regions;
A second step of measuring the luminance for each wavelength in each of the plurality of regions;
A third step of detecting light having a wavelength at which the luminance is outside the dynamic range of the imaging means;
A fourth step of correcting the brightness of the detected wavelength light;
And a fifth step of inspecting the surface of the image sensor based on the corrected captured image. The image sensor surface inspection method according to claim 1, wherein the wavelength is a wavelength of a complementary color of the image sensor. 3. The image sensor surface inspection method according to claim 1, wherein the fourth step is a step of correcting luminance of the detected wavelength light using a gradation filter of the detected wavelength. . 4. The image sensor surface according to claim 1, wherein the fourth step is a step of correcting the luminance of the light having the detected wavelength using the diode light source having the detected wavelength. 5. Inspection method. 4. The image sensor surface inspection method according to claim 3, wherein the fourth step is a step of using the gradation filter when the luminance is larger than a dynamic range of the imaging unit.