JP2019168388A - Image inspection method and image inspection device - Google Patents

Abstract

To provide an image inspection method and an image inspection device with which it is possible to reduce an effect of variations in brightness and hue of an image of a chip on a wafer and detect a defect with high accuracy.SOLUTION: An image inspection method includes: an image acquisition step of acquiring image data of a chip 41; a histogram calculation step of calculating a luminance histogram of the chip 41 on the basis of luminance of each pixel included in the acquired image data; a luminance threshold setting step of setting a luminance threshold that is a reference for defect discrimination to the chip 41 on the basis of a mode of the calculated luminance histogram; an extraction step of extracting, on the basis of the luminance threshold, binarized image data that includes a defect from the acquired image data of the chip 41; and a determination step of determining the defect with respect to the extracted binarized image data.SELECTED DRAWING: Figure 3

Description

  The present invention relates to an image inspection method and an image inspection apparatus, and more particularly to a technique of appearance image inspection for chips on a wafer.

  Conventionally, an appearance inspection of a chip surface on a wafer has been widely used as a sensor chip inspection process. As an inspection process of the sensor chip, an electrical inspection is performed, and a final appearance inspection is performed for the purpose of inspecting defects affecting the structural strength, sensor characteristics, and durability that cannot be detected by the electrical inspection.

  The appearance inspection of the chip on the wafer is performed using an image of the wafer. For example, image processing of a chip on a wafer is performed using an image inspection apparatus, and the appearance of the chip is inspected to determine the presence or absence of a defect. For chips for which the presence or absence of defects or the like could not be determined, visual inspection of images is performed, and inspection for presence or absence of defects is performed.

  In a conventional image inspection apparatus, a chip of about 2 mm square on a wafer is imaged with a coaxial illumination through a microscope. Since the sensing portion of the chip is hollow, luminance spots, that is, shading spots, are generated by reflection from the bottom surface. For this reason, when trying to detect a chip defect on the wafer by simple binarization, the shading due to the reflection of the illumination may be erroneously detected as a defect.

  For example, Patent Document 1 discloses a technique for detecting a defect of a chip on a wafer provided with a shading layer having a shading pattern and a covering layer covering the shading pattern of the shading layer. In Patent Document 1, the illumination light emitted from the illuminator has a wavelength at which the intensity of the light that is reflected or scattered from the coating layer and incident on the imaging device is larger than the light that is reflected from the grayscale layer and incident on the imaging device. Therefore, the influence of the shading pattern of the shading layer is reduced.

  Further, as a conventional image inspection method, there is a shading process for taking a difference from a reference image. However, the shape of the bottom surface of the hollow portion of the chip varies depending on the chip, and the position and the degree of luminance unevenness vary from chip to chip. In addition, since there is a difference in brightness and color difference between chips, it is difficult to inspect the chips on the wafer using the conventional shading process.

  As another conventional method, there is shading correction by Fourier transform, but it cannot be used when a large number of patterns such as heaters and temperature sensors are formed on the chip.

Japanese Patent Laid-Open No. 2018-11048

  Conventional image inspection methods are easily affected by variations in brightness and color in the chip image on the wafer, and it is difficult to realize defect inspection of the chip on the wafer with sufficient accuracy. Necessitated a visual inspection.

  The present invention has been made to solve the above-described problems, and is capable of detecting defects with higher accuracy by reducing the influence of variations in brightness and color in a chip image on a wafer. An object is to provide an inspection method and an image inspection apparatus.

  In order to solve the above-described problems, an image inspection method according to the present invention includes an image acquisition step of acquiring image data of a chip, and a histogram of the chip based on the luminance of each pixel included in the acquired image data. A first threshold value setting step for setting a first threshold value, which is a defect identification reference, in the chip based on the calculated mode value of the histogram, Based on a threshold value, a first extraction step for extracting a first region including a defect from the acquired image data of the chip, and a determination for determining a defect with respect to the extracted first region And a step.

  Further, in the image inspection method according to the present invention, a conversion step of converting the acquired image data into hue image data indicating the hue component, and a second step set in advance as a hue component indicating the defect of the chip. Based on the threshold value, the first area extracted from the hue image data in the second extraction step for extracting a second area including a defect, the first extraction step, and the second extraction step, respectively. And a synthesis step of generating a synthesis region by synthesizing the second region and the second region, and the determination step may determine a defect in the synthesis region.

  In the image inspection method according to the present invention, the determination step may determine the defect with reference to information on the defect stored in the storage unit.

  The image inspection method according to the present invention includes an image acquisition step of acquiring image data of a chip, a conversion step of converting the acquired image data into hue image data indicating the hue component, and a defect of the chip A second extraction step for extracting a second area including a defect from the hue image data based on a second threshold value set in advance as a hue component indicating the color component; and for the extracted second area And a determination step for determining a defect.

  An image inspection apparatus according to the present invention includes an image acquisition unit that acquires image data of a chip, a histogram calculation unit that calculates a histogram of the chip based on luminance of each pixel included in the acquired image data, Based on the calculated mode value of the histogram, based on the first threshold value setting unit for setting a first threshold value that is a defect identification standard in the chip, and on the first threshold value, A first extraction unit that extracts a first region including a defect from the acquired image data of the chip, and a determination unit that performs defect determination on the extracted first region. And

  In the image inspection apparatus according to the present invention, a conversion unit that converts the acquired image data into hue image data indicating the hue component, and a second threshold set in advance as a hue component indicating the defect of the chip. Based on the value, the second extraction unit that extracts a second region including a defect from the hue image data, the first extraction unit, and the first extraction unit respectively extracted by the second extraction unit and the second extraction unit The image forming apparatus may further include a combining unit that combines the second region and generates a combined region, and the determination unit may perform defect determination on the combined region.

  An image inspection program according to the present invention causes a computer to execute the image inspection method.

  According to the present invention, the histogram of the chip is calculated based on the luminance of each pixel included in the image data of the chip, and the first threshold value that is the defect identification reference is set. Defects can be detected with higher accuracy by reducing the influence of variations in brightness and color.

FIG. 1 is a block diagram showing a configuration of an image inspection apparatus according to the first embodiment of the present invention. FIG. 2 is a block diagram showing a hardware configuration of the image inspection apparatus according to the first embodiment of the present invention. FIG. 3 is a flowchart for explaining the operation of the image inspection apparatus according to the first embodiment of the present invention. FIG. 4 is a flowchart of luminance image processing according to the first embodiment of the present invention. FIG. 5 is an explanatory diagram of the luminance threshold value according to the first embodiment of the present invention. FIG. 6 is a block diagram showing a configuration of an image inspection apparatus according to the second embodiment of the present invention. FIG. 7 is a flowchart for explaining the operation of the image inspection apparatus according to the second embodiment of the present invention. FIG. 8 is a flowchart of hue image processing according to the second embodiment of the present invention. FIG. 9 is an explanatory diagram of hue image processing according to the second embodiment of the present invention. FIG. 10 is an explanatory diagram of hue image processing according to the second embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to FIGS.
[First Embodiment]
FIG. 1 is a block diagram showing a configuration of an image inspection apparatus 1 according to the first embodiment of the present invention.
The image inspection apparatus 1 determines the defect of the chip 41 based on the image of the chip 41 on the inspection target wafer 4 placed on the table 3 captured by the camera 2.

  The camera 2 is connected to the image inspection apparatus 1 and captures an optical image of the chip 41 on the wafer 4 to be inspected. The camera 2 uses an image sensor such as a two-dimensional CCD, and converts an optical image formed on a light receiving surface (not shown) into an electrical signal. For example, the camera 2 is provided with a color filter (not shown) having a Bayer arrangement in which three primary colors of red (R), green (G), and blue (B) are arranged, and images light that has passed through the color filter. In the present embodiment, the image signal output from the camera 2 indicates color image data.

  For example, the camera 2 captures images by dividing the chip 41 on the wafer 4 into a plurality of areas.

On the table 3, a wafer 4 to be inspected is installed.
For example, a chip 41 such as a microflow sensor is formed on the wafer 4. The chip 41 has an insulating film layer formed on the surface of a fine silicon chip, a hollow portion such as a diaphragm portion is formed at the center, and a sensor such as a temperature sensor is formed of a platinum thin film.

  The chip 41 on the wafer 4 may include a defect that affects strength such as a crack, a defect that affects sensor characteristics such as a pattern abnormality, and a defect that affects durability such as Pt exposure. The image inspection apparatus 1 detects these defects and determines whether the detected defect is a crack or a pattern abnormality.

Next, each configuration of the image inspection apparatus 1 will be described.
The image inspection apparatus 1 includes an image acquisition unit 10, a luminance image processing unit 11, a storage unit 12, a determination unit 13, and a display unit 14.

  The image acquisition unit 10 acquires a color image signal of the chip 41 on the wafer 4 captured by the camera 2. More specifically, the image acquisition unit 10 acquires a color image signal for each of a plurality of areas in each of the plurality of chips 41. The acquired image data of the chip 41 is stored in the storage unit 12.

The luminance image processing unit 11 includes a luminance histogram calculation unit 110, a luminance threshold value setting unit (first threshold value setting unit) 111, a gray scale image generation unit 112, and an extraction unit (first extraction unit) 113.
The luminance image processing unit 11 performs image processing on the image data of the chip 41 acquired by the image acquisition unit 10 based on the luminance, and generates binarized image data of a region including a defect.

  The luminance histogram calculation unit 110 reads the image data of the chip 41 from the storage unit 12, and calculates a histogram based on the luminance of each pixel included in the image data for each image data of the chip 41. More specifically, the luminance histogram calculation unit 110 calculates a histogram representing the luminance distribution of the image for each image data of the chip 41, and obtains a mode value for each histogram. The calculated histogram is stored in the storage unit 12 in association with the image data and the mode value.

  The luminance threshold value setting unit 111 is based on the mode value of the histogram calculated by the luminance histogram calculation unit 110, and is a luminance threshold value (first threshold value) that is a defect identification reference in the chip 41 on the wafer 4. ) Is set for each image data of the chip 41. The luminance threshold is set so as to be linked to the mode value. The luminance threshold value is stored in the storage unit 12 in association with the image data.

  The gray scale image generation unit 112 converts the color image data of the chip 41 into gray scale image data representing 256 shades of black and white. The converted gray scale image data is stored in the storage unit 12 in association with the corresponding color image data.

  The extraction unit 113 extracts a region including a defect from the grayscale image data of the chip 41 obtained by the grayscale image generation unit 112 based on the luminance threshold value. More specifically, the extraction unit 113 divides a region brighter than a luminance threshold value and a dark region using the luminance threshold value set by the luminance threshold value setting unit 111, and binary values of white and black Convert to converted image data. In this case, binarized image data (first area) is extracted in which gray areas are all black in areas that are darker than the luminance threshold and white are areas that are brighter than the luminance threshold. The binarized image data for each extracted image data is stored in the storage unit 12.

  The extraction unit 113 extracts HSV (hue (H), saturation (S), brightness (V) from RGB color image data instead of extracting binarized image data based on luminance from grayscale image data. )) After converting into image data in a color space, binarized image data may be extracted based on brightness (V).

  The storage unit 12 stores a luminance histogram, a luminance threshold, grayscale image data, binarized image data, and the like for each color image data of the chip 41.

  The determination unit 13 determines a defect based on the binarized image data obtained for each image data of the chip 41 by the luminance image processing unit 11. More specifically, the determination unit 13 refers to defect data that has been classified and defined in advance, and determines which defect the binarized image data corresponds to.

  The display unit 14 displays the binarized image data extracted by the extraction unit 113 and the determination result by the determination unit 13.

[Hardware configuration of image inspection device]
Next, the hardware configuration of the image inspection apparatus 1 according to the present embodiment will be described with reference to FIG.

  As shown in FIG. 2, the image inspection apparatus 1 includes an arithmetic device 102 having a CPU 103 and a main storage device 104 connected via a bus 101, a communication control device 105, an external storage device 106, a display device 107, an I / O device. It can be realized by a computer provided with F108, camera 109, etc., and a program for controlling these hardware resources.

  The CPU 103 and the main storage device 104 constitute an arithmetic device 102. A program for the CPU 103 to perform various controls and calculations is stored in the main storage device 104 in advance.

  The communication control device 105 is a control device for network connection between the image inspection device 1 and various external electronic devices. In the present embodiment, the image inspection apparatus 1 may be connected to a management server (not shown) via a network.

  The external storage device 106 includes a readable / writable storage medium and a drive device for reading / writing various information such as programs and data from / to the storage medium. The external storage device 106 can use a semiconductor memory such as a hard disk or a flash memory as a storage medium. The external storage device 106 is an image information storage unit 106a, a defect information storage unit 106b, a program storage unit 106c, and other storage devices not shown. For example, the external storage device 106 backs up programs and data stored in the external storage device 106. A storage device or the like can be included.

The image information storage unit 106a corresponds to the storage unit 12 described in FIG.
The defect information storage unit 106b stores information related to defects that are defined in advance.

  The program storage unit 106c executes processing necessary for image inspection such as luminance histogram calculation processing, luminance threshold setting processing, image conversion processing, binarized image data extraction processing and determination processing in the present embodiment. Various programs are stored.

  The display device 107 is configured by a liquid crystal display or the like. The display unit described with reference to FIG.

The I / F 108 is an interface for connecting various devices. In the present embodiment, the camera 109 is connected, and the image data of the chip 41 is acquired via the I / F 108.
The camera 109 corresponds to the camera 2 described in FIG.

[Operation of image inspection device]
Next, an operation when the image inspection apparatus 1 having the above-described configuration inspects a defect of the chip 41 on the wafer 4 will be described with reference to a flowchart of FIG. First, the image acquisition unit 10 of the image inspection apparatus 1 acquires the image data of the chip 41 on the wafer 4 captured by the camera 2 (step S1).

  Next, the luminance image processing unit 11 performs image processing based on luminance for each image data acquired by the image acquisition unit 10 (step S2).

Here, the luminance image processing by the luminance image processing unit 11 will be described with reference to FIG.
The luminance histogram calculation unit 110 calculates a histogram of the color image data of the chip 41 acquired by the image acquisition unit 10 (step S120). As described above, the luminance histogram calculation unit 110 calculates a luminance histogram for each image data of the chip 41.

  Next, the brightness threshold value setting unit 111 obtains the mode value in the brightness histogram for each image data of the chip 41 calculated in step S120, and based on the mode value, the brightness threshold value serving as a defect identification reference is obtained. A threshold value is set for each luminance histogram (step S121).

Thereafter, the gray scale image generation unit 112 converts the color image data of the chip 41 acquired by the image acquisition unit 10 into gray scale image data (step S122).
Next, the extraction unit 113 extracts binary image data of grayscale image data corresponding to the color image data of the chip 41 based on the luminance threshold value set in step S121.

  As shown in FIGS. 5A and 5B, in the luminance histogram calculated for each color image data of the chip 41, mode value = 102 in (a) and mode value = 95 in (b). Is obtained. Based on these mode values, luminance threshold values serving as defect identification criteria are set. In FIG. 5A, the luminance threshold value = 88, whereas in FIG. 5B, the luminance threshold value = 81, which is a different luminance threshold value for each image data.

  Next, in the flowchart of FIG. 3, the determination unit 13 refers to the defect information defined in advance for the binarized image data extracted for each image data, and the defect indicated by the binarized image data It is determined whether the defect belongs to the classification (step S3). Finally, the display unit 14 displays the determination result on the screen (step S4).

  As described above, according to the first embodiment, the brightness threshold value which becomes the defect identification reference is set for each image data of the chip 41 on the wafer 4, so that different brightness and color for each image data. More accurate defect detection with reduced influence of taste variation is realized. That is, since the image inspection is performed by setting the brightness threshold value linked to the variation in brightness and color for each image data, it is possible to improve false detection and oversight.

[Second Embodiment]
Next, a second embodiment of the present invention will be described. In the following description, the same components as those in the first embodiment described above are denoted by the same reference numerals, and the description thereof is omitted.

  In the first embodiment, the image inspection apparatus 1 includes a luminance image processing unit 11, sets a luminance threshold value for each image data of the chip 41 on the wafer 4, and extracts binarized image data. Explained when to do. On the other hand, in the second embodiment, the image inspection apparatus 1A is different from the image inspection apparatus 1 according to the first embodiment in that it further includes a hue image processing unit 15 and a combining unit (combining unit) 16. Is different.

[Outline of Image Inspection Apparatus According to Second Embodiment]
First, an outline of the image inspection apparatus 1A according to the second embodiment will be described.
In the chip 41 on the wafer 4, for example, a region where a diaphragm is formed is in a hollow state. Such a region is engraved by wet etching. However, for example, when coaxial epi-illumination is irradiated, luminance unevenness occurs due to reflection from the bottom surface of the diaphragm portion.

  In addition, since the shape of the bottom surface varies depending on the chip 41 and the position and level of the brightness unevenness are different, even if the difference detection with the reference image, which is a conventionally used technique, is detected, the brightness unevenness is detected together with the defect detection. May end up.

  The image inspection apparatus 1A according to the present embodiment pays attention to the feature that the hue does not change even if the brightness of the image is different from the fact that the luminance unevenness due to the reflection of the illumination is erroneously detected as a defect. Then, the binarized image data is extracted from the hue image data indicating the hue in the image of the chip 41.

[Configuration of image inspection device]
FIG. 6 is a block diagram showing a configuration of an image inspection apparatus 1A according to the second embodiment.
Hereinafter, the image inspection apparatus 1A will be described focusing on a configuration different from that of the image inspection apparatus 1 according to the first embodiment.

The hue image processing unit 15 includes a hue image generation unit (conversion unit) 150, a hue threshold value setting unit (second threshold value setting unit) 151, and an extraction unit (second extraction unit) 152.
The hue image processing unit 15 converts binarized image data (second image data) from image data of the chip 41 on the wafer 4 based on a hue threshold value (second threshold value) corresponding to a preset hue range. Region).

  The hue image generation unit 150 converts the color image data of the chip 41 acquired by the image acquisition unit 10 into hue image data of 256 gradations. More specifically, the hue image generation unit 150 converts the color image into hue (H) image data in the HSV color space. The obtained hue image data is stored in the storage unit 12.

  The hue threshold value setting unit 151 sets a hue threshold value corresponding to a predetermined desired hue for each image data. The set hue threshold value is stored in the storage unit 12. For example, a hue threshold value corresponding to a specific defect classified in advance, such as Pt exposure, is set.

  The extraction unit 152 extracts binarized image data from the hue image data of the chip 41 obtained by the conversion by the hue image generation unit 150. The extracted binarized image data is stored in the storage unit 12 in association with the image data.

  The combining unit 16 is extracted by the binarized image data based on the luminance threshold set for each image data extracted by the extraction unit 113 of the luminance image processing unit 11 and the extraction unit 152 of the hue image processing unit 15. The binarized image data based on the hue threshold value set for each image data is combined to generate combined binarized image data (composite region). The generated composite binarized image data is stored in the storage unit 12 in association with the image data.

  More specifically, the combining unit 16 includes the binarized image data extracted by the extracting unit 113 of the luminance image processing unit 11 and the binarized image data extracted by the extracting unit 152 of the hue image processing unit 15. A sum area (OR: superposition) is taken and generated as synthesized binary image data indicating a defective area.

  Note that the binarized image data extracted by the extraction unit 113 of the luminance image processing unit 11 is converted from RGB color image data to grayscale image data or HSV color space image data, and the brightness (V) therein. Can handle.

[Operation of image inspection device]
Next, the operation of the image inspection apparatus 1A will be described with reference to the flowcharts of FIGS.
First, the image acquisition unit 10 acquires image data of the chip 41 on the wafer 4 captured by the camera 2 (step S20). Next, the luminance image processing unit 11 performs luminance image processing for each image data, and extracts binarized image data from the image data based on the luminance threshold value set for each image data (step S21). .

  Next, the hue image processing unit 15 performs hue image processing on the image data of the chip 41 acquired by the image acquisition unit 10 (step S22).

  As illustrated in FIG. 8, the hue image generation unit 150 converts the color image data of the chip 41 acquired by the image acquisition unit 10 into hue image data (step S220). Next, the hue threshold value setting unit 151 sets a hue threshold value corresponding to a preset desired hue (step S221).

  Thereafter, the extraction unit 152 extracts the binarized image data based on the set hue threshold value in the hue image data of the chip 41 obtained by the conversion by the hue image generation unit 150 (step S222). .

  Here, as shown in FIG. 9, the image data of the area of the chip 41 acquired as the color image data is converted into the hue image data, and is further converted into white and black based on the set hue threshold value. The binarized image data configured is extracted.

  Further, as shown in FIG. 10, in the histogram showing the hue distribution of the area of the chip 41, a hue threshold corresponding to the range of red to yellow is set as a desired detection range, so that the target defect is detected. It becomes possible to extract.

  Next, as illustrated in FIG. 7, the combining unit 16 includes the binarized image data extracted by the extraction unit 113 of the luminance image processing unit 11 and the binary extracted by the extraction unit 152 of the hue image processing unit 15. The combined image data is combined to generate combined binary image data (step S23).

  After that, the determination unit 13 refers to information relating to the defects defined in advance for the synthesized binary image data in which the results of the luminance image processing unit 11 and the hue image processing unit 15 are combined by the combining unit 16. Defect determination is performed (step S24).

  And the display part 14 displays the determination result of the defect by the determination part 13 on a display screen (step S25).

  As described above, according to the second embodiment, the image inspection apparatus 1A extracts the binarized image data from the hue image data of the image data based on the hue threshold value corresponding to the desired hue. Therefore, it is possible to suppress erroneous detection of luminance unevenness as a defect and improve the detection accuracy of the defect.

  Further, according to the second embodiment, the binarized image data using the luminance threshold value set for each image data, the binarized image data of the hue image data using the hue threshold value, and Is synthesized. Thereby, it is possible to more clearly detect the entire shape of the defect due to the foreign matter with less change in background and luminance. As a result, it is possible to sufficiently improve the defect detection accuracy and eliminate the need for visual appearance inspection.

  Although the embodiments of the image inspection method and the image inspection apparatus of the present invention have been described above, the present invention is not limited to the described embodiments, and those skilled in the art can assume the scope of the invention described in the claims. Various possible modifications can be made.

  For example, in the embodiment described above, the case where color image data using RGB is used has been described. However, the color image data may be based on CMYK (cyan, magenta, yellow, black), for example.

  DESCRIPTION OF SYMBOLS 1, 1A ... Image inspection apparatus, 2 ... Camera, 3 ... Table, 4 ... Wafer, 41 ... Chip, 10 ... Image acquisition part, 11 ... Luminance image processing part, 12 ... Memory | storage part, 13 ... Determination part, 14 ... Display , 15 ... Hue image processing unit, 16 ... Coupling unit, 101 ... Bus, 102 ... Arithmetic device, 103 ... CPU, 104 ... Main storage device, 105 ... Communication control device, 106 ... External storage device, 106a ... Image information storage 106b ... defect information storage unit 106c ... program storage unit 107 ... display device 108 ... I / F 109 ... camera 110 110 luminance histogram calculation unit 111 luminance threshold setting unit 112 gray scale Image generation unit 113, 152 ... extraction unit, 150 ... hue image generation unit, 151 ... hue threshold value setting unit.

Claims (7)

An image acquisition step of acquiring image data of the chip;
A histogram calculation step of calculating a histogram of the chip based on the luminance of each pixel included in the acquired image data;
A first threshold value setting step for setting a first threshold value, which is a defect identification reference, in the chip based on the calculated mode value of the histogram;
A first extraction step of extracting a first region including a defect from the acquired image data of the chip based on the first threshold;
An image inspection method comprising: a determination step of determining a defect with respect to the extracted first region. The image inspection method according to claim 1,
A conversion step of converting the acquired image data into hue image data indicating the hue component;
A second extraction step of extracting a second region including a defect from the hue image data based on a second threshold value set in advance as a hue component indicating the defect of the chip;
A combining step of generating a combined region by combining the first region and the second region respectively extracted in the first extracting step and the second extracting step;
Further comprising
In the image inspection method, the determination step includes determining a defect in the composite region. The image inspection method according to claim 1 or 2,
An image inspection method according to claim 1, wherein in the determination step, the defect is determined with reference to information on the defect stored in the storage unit. An image acquisition step of acquiring image data of the chip;
A conversion step of converting the acquired image data into hue image data indicating the hue component;
A second extraction step of extracting a second region including a defect from the hue image data based on a second threshold value set in advance as a hue component indicating the defect of the chip;
An image inspection method comprising: a determination step of determining a defect with respect to the extracted second region. An image acquisition unit for acquiring image data of the chip;
A histogram calculation unit that calculates a histogram of the chip based on the luminance of each pixel included in the acquired image data;
A first threshold value setting unit for setting a first threshold value, which is a defect identification reference, in the chip based on the calculated mode value of the histogram;
A first extraction unit that extracts a first region including a defect from the acquired image data of the chip based on the first threshold;
An image inspection apparatus comprising: a determination unit configured to determine a defect with respect to the extracted first region. The image inspection apparatus according to claim 5,
A conversion unit that converts the acquired image data into hue image data indicating the hue component;
A second extraction unit that extracts a second region including a defect from the hue image data based on a second threshold value preset as a hue component indicating a defect of the chip;
A synthesis unit that generates a synthesis region by synthesizing the first region and the second region extracted by the first extraction unit and the second extraction unit;
The determination unit performs a defect determination on the composite region.   An image inspection program causing a computer to execute the image inspection method according to claim 1.