JP2008175550A - Flaw detector and flaw detection method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a flaw detector capable of reducing the misdetection of a flaw, and to provide a flaw detection method. <P>SOLUTION: A transformation part 24 for time delay coordinate system transforms data comprising the pixel values of respective pixels which constitute the image of an inspection target having a cycle pattern formed thereto to the data of a time delay coordinate system of a dimension (m) (m is an integer of 2 or above). A flaw detection part 25 judges whether the position showing the data of the time delay coordinate system is contained in a predetermined region on an (m) dimensional space to detect the flaw on the basis of the judge result. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a defect detection apparatus and a defect detection method for detecting a defect on an inspection object using an image obtained by imaging an inspection object on which a periodic pattern is formed.

  Conventionally, a periodic pattern of an electronic circuit formed on an inspection object (such as a substrate for a flat panel display) is sequentially read as an image by an imaging device such as a line sensor camera, and the image is used on the inspection object. A pattern inspection for detecting a defect is performed. As a pattern inspection technique, the brightness values of patterns shifted by one period on the image are compared, and when the difference in brightness value exceeds a predetermined value, the corresponding portion is detected as a defect, so-called adjacent Comparison methods are known.

When a line sensor camera is used, imaging signals are output in units of columns (lines) from light receiving elements arranged in a row. In the adjacent comparison method, the data for one cycle of the pattern is sequentially compared with the data for one cycle corresponding to the same position while storing the data for one column in the storage unit (for example, Patent Document 1). reference).
JP 2006-78421 A

  However, in the conventional pattern inspection by the adjacent comparison method, when the pattern width fluctuates by about several pixels of the detection resolution due to fluctuations in the manufacturing process, etc., the fluctuation (fluctuation) of the pattern width is within the allowable range for manufacturing. However, there was a problem that it was erroneously detected as a defect. Hereinafter, this specific example will be described.

  For example, a solid line 901 in FIG. 9 schematically shows a part of a change in luminance signal when a pattern without fluctuations in the manufacturing process is imaged, and a broken line 902 is at a position adjacent to the pattern without fluctuations. A part of change of the luminance signal is schematically shown when a pattern whose period is gradually shortened due to the occurrence of fluctuation is captured. FIG. 10 shows the luminance difference between the luminance signal indicated by the solid line 901 and the luminance signal indicated by the broken line 902 by the adjacent comparison. As shown in FIG. 10, the brightness difference gradually increases as it progresses in the right direction in the figure. For example, when the threshold value for defect determination is 1.0, the pattern after the fifth cycle is detected as a defect. .

  The present invention has been made in view of the above-described problems, and an object of the present invention is to provide a defect detection apparatus and a defect detection method that can reduce erroneous detection of defects.

  The present invention has been made in order to solve the above-described problem, and data consisting of pixel values of pixels constituting an image of an inspection object on which a periodic pattern is formed is represented by a dimension m (m is an integer of 2 or more). Data conversion means for converting the data into the time delay coordinate system, and whether or not the position indicated by the data in the time delay coordinate system in the m-dimensional space is included in the predetermined area, and the defect is determined based on the determination result A defect detection apparatus comprising defect detection means for detecting.

  The present invention also includes a step of converting data composed of pixel values of each pixel constituting an image of the inspection object on which a periodic pattern is formed into data of a time delay coordinate system of dimension m (m is an integer of 2 or more). And a step of determining whether or not a position indicated by the data of the time delay coordinate system in the m-dimensional space is included in a predetermined region, and detecting a defect based on the determination result It is a detection method.

  Due to the characteristics of the data in the time delay coordinate system, if the inspection object has no defects and the periodic relationship between the data before conversion to the time delay coordinate system is maintained, the dimensions of the periodic pattern will vary. Even if there is, the data fits within a predetermined area on the m-dimensional space of the time delay coordinate system. On the other hand, if there is a defect on the inspection object and a part of the periodic relationship between the data before the conversion to the time delay coordinate system is disturbed, the above-described problem occurs in the m-dimensional space of the time delay coordinate system. Data that does not fit in the area is generated.

  According to the present invention, in the m-dimensional space, it is determined whether or not the data of the time delay coordinate system is included in the predetermined region, and the defect is detected based on the determination result, thereby reducing the erroneous detection of the defect. The effect of being able to be obtained.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 shows the configuration of a pattern inspection system according to an embodiment of the present invention. The imaging device 1 includes a line sensor camera or the like, and images an inspection object or a non-defective product having a periodic pattern formed on the surface, and generates an imaging signal. The image processing device 2 (corresponding to the defect detection device of the present invention) processes an image based on the imaging signal input from the imaging device 1 and detects a defect on the inspection object. The display device 3 displays the result of the defect detection process performed by the image processing device 2.

  Hereinafter, the internal configuration of the image processing apparatus 2 will be described. The image input unit 21 generates image data by A / D converting the imaging signal input from the imaging device 1. The storage unit 22 stores image data, data necessary for defect detection processing, result data of defect detection processing, and the like. The control unit 23 controls each unit in the image processing apparatus 2.

  The time delay coordinate system conversion unit 24 considers a data sequence of pixel values of pixels arranged in a line in an image as time series data, and converts each data constituting the data sequence to a time delay known in the field of nonlinear analysis. Convert to coordinate system (Kazuyuki Aihara, “Basics and Applications of Chaos Time Series Analysis”, Sangyo Tosho, 2000). The defect detection unit 25 performs a process of detecting a defect using the data converted into the time delay coordinate system. The data output unit 26 outputs the result of the defect detection process and the like to the display device 3.

  Next, a defect detection method in this embodiment will be described. When an imaging signal obtained by imaging the inspection object is input to the image processing apparatus 2, the image input unit 21 generates image data and stores it in the storage unit 22. The time delay coordinate system conversion unit 24 reads the image data from the storage unit 22, and converts the data composed of the pixel values of each pixel constituting the image into a delay time τ (τ is an integer of 1 or more), dimension m (m is 2 or more) To the data of the time delay coordinate system.

  The delay time τ indicates the number of pixels in this embodiment. Τ may be determined from the pattern period (for example, the design value of the pattern period) when there is no variation (fluctuation) in the pattern width, τ may be determined by a known method such as the use of autocorrelation, τ may be obtained experimentally. Further, the number m may be any number, but usually 2 or 3 may be selected.

  As an example, it is assumed that the time delay coordinate system conversion unit 24 reads out the pixel values of 50 pixels arranged in a line on the image from the storage unit 22. FIG. 2A shows the pixel values of these 50 pixels. The pixel value of each pixel takes any value from 0 to 255. The array for storing the pixel values of these 50 pixels is F, and the t-th data is F (t) (t = 1, 2,..., 50).

The time delay coordinate system conversion unit 24 considers the data of the array F as time series data that advances one pixel in one unit time in order from the t = 1 pixel, and converts the data in the time delay coordinate system of the delay time τ and dimension m. Convert. A general conversion to the array G of the time delay coordinate system is expressed by the following equation (1).
G (t) = (F (t), F (t + τ), F (t + 2τ),..., F (t + (m−1) τ)) (1)

In this example, it is assumed that the delay time τ = 1 and the dimension m = 2. In the case of this example, the conversion to the array G of the time delay coordinate system is expressed by the following equation (2).
G (t) = (F (t), F (t + 1)) (2)

  FIG. 2B shows each element value of G (t). X indicates the value of F (t), and Y indicates the value of F (t + 1). By the conversion of equation (2), the position values indicated by the black and white circles in FIG. 2A are converted to the position values indicated by the black and white circles in FIG.

  Hereinafter, the principle of the defect detection process will be described. When the element value F (t) of G (t) is regarded as the X coordinate value and the element value F (t + 1) is regarded as the Y coordinate value, each data of G (t) is plotted in the m = 2 dimensional space as shown in FIG. become.

  If there is no defect on the inspection object due to the characteristics of the data in the time delay coordinate system and the periodic relationship between the data in the array F is maintained, the time can be changed even if the dimension of the periodic pattern varies. On the m-dimensional space of the delayed coordinate system, the position indicated by the data of the array G is within a predetermined area. On the other hand, if there is a defect on the inspection object and a part of the periodic relationship between the data of the array F is disturbed, the m-dimensional of the time delay coordinate system among the positions indicated by the data of the array G A position that does not fit within a predetermined area occurs in space.

  In array F, F (7) is an abnormal value, and the other values are normal values. In FIG. 3, G (6) and G (7) including the abnormal value F (7) are respectively plotted at positions C and D different from regions A and B where other normal values are plotted. In this example, the delay time τ = 1, the dimension m = 2, and the pixel value of the defective pixel is not related to the pixel value of the adjacent pixel, so that the difference in the plot position as described above. Appears. By the way, when the delay time τ = 1 and the dimension m = 3, plotting is performed with a ternary relationship between the pixel value of the target pixel, the pixel value of the adjacent pixel, and the pixel value of the adjacent pixel. The position will be determined.

  From the above, G (6) and G (7) plotted at positions distant from the regions A and B where normal values are plotted on the two-dimensional space as shown in FIG. 3 are used as defect candidate data. It is possible to detect that F (7) common to the element values constituting the data is an abnormal value. From this detection result, it is possible to determine that the seventh pixel is a defective pixel.

  In the present embodiment, the positions and extents of areas A and B including points in a two-dimensional space in which only the pixel values of normal pixels not including the pixel values of defective pixels are set as X coordinate values and Y coordinate values are determined in advance. (The determination method will be described later). Depending on the degree of fluctuation (fluctuation) of the pattern width, the distribution of points in the two-dimensional space (hereinafter referred to as normal points) in which only the pixel values of normal pixels are X coordinate values and Y coordinate values changes. Therefore, even if there is some variation (fluctuation) in the pattern width, the areas A and B are allowed to have a certain extent so that the normal points are included in the areas A and B. Hereinafter, areas such as areas A and B that should include normal points are referred to as fluctuation allowable areas.

Next, a specific procedure of defect detection processing according to the above-described principle will be described. First, a procedure for generating data of the array H indicating the distribution of the fluctuation allowable area in the m-dimensional space of the delay time coordinate system (that is, the procedure for determining the fluctuation allowable area) will be described. Assuming that the pixel value takes any value from 0 to 255, the array H is composed of 256 ^m data corresponding to all coordinate positions in the m-dimensional space of the delay time coordinate system. It is assumed that the data of the array H corresponding to the coordinate position in the fluctuation allowable area is 1, and the data of the array H corresponding to the coordinate position outside the fluctuation allowable area is 0.

  When the periodic pattern on the inspection object is known in advance, the fluctuation allowable region may be obtained by calculation based on the design value or may be obtained from actual data. Hereinafter, a method for generating data of the array H from actual data will be described with reference to FIG.

  After the start of processing, the time-delayed coordinate system conversion unit 24 first initializes the value of the variable i for specifying the processing target data (i = 1) (step S100). In addition, the time delay coordinate system conversion unit 24 initializes all the data of the array H stored in the storage unit 22 (step S110). Subsequently, the time delay coordinate system conversion unit 24 reads data of m pixels necessary for processing from the storage unit 22 as described below (step S120).

  That is, the time-delay coordinate system conversion unit 24, for each value of n = 1 to m (where n is an integer), the pixel of the pixel that is separated from the target pixel number i by (n−1) × τ The value is read from the storage unit 22, and each pixel value is stored in the i-th, (i + τ) -th,..., (I + (m−1) × τ) -th array F, respectively. For example, when m = 2 and i = 0, the time delay coordinate system conversion unit 24 reads out the pixel values of the i-th and (i + τ) -th pixels from the storage unit 22, and sets each pixel value to F (i), Let F (i + τ). The image data used for the processing is obtained by imaging a reference object such as a non-defective non-defective substrate.

  Subsequently, the time delay coordinate system conversion unit 24 converts the data read from the storage unit 22 into m-dimensional time delay coordinate system data (step S130). That is, similarly to the equation (1), the time delay coordinate system conversion unit 24 generates G (i) that is the i-th data of the array G of the time delay coordinate system from the data of the array F.

  The data of the arrays F and G and the value of the number i are transferred from the time delay coordinate system conversion unit 24 to the defect detection unit 25. Further, the defect detection unit 25 reads the data of the array H from the storage unit 22. The defect detection unit 25 refers to G (i) which is the i-th data of the array G, and points within a predetermined range centered on the point indicated by G (i) in the m-dimensional space (for example, G (i) 1 is stored in the data of the array H corresponding to each of the points within a few pixels from the point indicated by () (step S140). At this time, if the data of the array H is already 1, the defect detection unit 25 does not update the data (or may update it to 1), and proceeds to processing of the next data.

  Subsequently, the defect detection unit 25 determines whether or not the variable i has the same value as the predetermined maximum number of pixels N (step S150). If i and N do not match, the defect detection unit 25 notifies the time delay coordinate system conversion unit 24 of the next data processing. Receiving the notification, the time delay coordinate system conversion unit 24 adds 1 to the value of i (step S160), and executes the process of step S120 again. The above process is repeated until i becomes the same value as N.

  On the other hand, when i and N match in step S150, the defect detection unit 25 stores the data of the array H in the storage unit 22 (step S170). This completes the process of determining the fluctuation allowable area. As a result of the above processing, the data corresponding to the coordinate position within the fluctuation allowable area is 1 in the data of the array H, and the data corresponding to the coordinate position outside the fluctuation allowable area is 0.

  Next, the procedure of defect detection processing will be described with reference to FIG. After the defect detection process is started, the time delay coordinate system conversion unit 24 first initializes the value of the variable i for specifying the data to be processed (i = 1) (step S200). Subsequently, the time delay coordinate system conversion unit 24 reads data of m pixels necessary for processing from the storage unit 22 in the same manner as in step S110 of FIG. 4 (step S210).

  Subsequently, the time delay coordinate system conversion unit 24 converts the data read from the storage unit 22 into m-dimensional time delay coordinate system data (step S220). That is, similarly to the equation (1), the time delay coordinate system conversion unit 24 generates G (i) that is the i-th data of the array G of the time delay coordinate system from the data of the array F.

  The data of the arrays F and G and the value of the number i are transferred from the time delay coordinate system conversion unit 24 to the defect detection unit 25. Further, the defect detection unit 25 reads the data of the array H from the storage unit 22. The defect detection unit 25 refers to G (i) that is the i-th data of the array G, and refers to the data of the array H corresponding to the point indicated by G (i) in the m-dimensional space (step S230).

  Subsequently, the defect detection unit 25 determines whether or not the data of the array H referred to in step S230 is 1 (step S240). If the data of the array H referred to in step S230 is 1, the point on the m-dimensional space indicated by G (i) is in the fluctuation allowable region. If the data of the array H referenced in step S230 is 0, the point on the m-dimensional space indicated by G (i) is outside the fluctuation allowable region.

  If the point on the m-dimensional space indicated by G (i) is within the fluctuation allowable region, the process proceeds to step S260. If the point in the m-dimensional space indicated by G (i) is outside the fluctuation allowable region, the defect detection unit 25 determines that G (i) is defect candidate data (step S250). . Subsequently, the process proceeds to step S260.

  The defect detection unit 25 determines whether or not the variable i has the same value as the predetermined maximum pixel number N (step S260). If i and N do not match, the defect detection unit 25 notifies the time delay coordinate system conversion unit 24 of the next data processing. Receiving the notification, the time delay coordinate system conversion unit 24 adds 1 to the value of i (step S270) and executes the process of step S210 again. The above process is repeated until i becomes the same value as N.

  On the other hand, when i and N match in step S260, the defect detection unit 25 specifies data of the array F including the pixel values of the defective pixels from the defect candidate data of the array G obtained in step S250. Then, the position of the defective pixel is specified (step S280). When the defect detection unit 25 stores the result of the defect detection process in the storage unit 22 (step S290), the defect detection process ends.

  Next, a procedure of a modification of the defect detection process in the present embodiment will be described. First, the procedure for determining the fluctuation allowable area will be described with reference to FIG. The processes in steps S300 to S330 and S350 to S370 in FIG. 6 are the same as the processes in steps S100 to S130 and S150 to S170 in FIG.

  In step S340, the defect detection unit 25 refers to G (i) that is the i-th data of the array G, and points within a predetermined range centered on the point in the m-dimensional space indicated by G (i) (for example, 1 is added to the data of the array H corresponding to each of the points within a few pixels from the point indicated by G (i). When the process of step S340 is repeated until i and N match, a large value is stored in the data of the array H corresponding to the data of the array G composed of only normal pixel values among the data of the array G. The

  In step S380, the defect detection unit 25 calculates the defect determination threshold D used for the determination process when extracting defect candidate data in the defect detection process from the data of the array H, and stores the defect determination threshold D in the storage unit 22. To do. In order to obtain the defect determination threshold D, for example, the average value μ and the standard deviation σ of the data of the array H excluding 0 may be obtained and D = μ + 3σ. This completes the process of determining the fluctuation allowable area.

  As a result of the above processing, the data corresponding to the coordinate position within the fluctuation allowable area in the data of the array H is a value equal to or greater than the defect determination threshold, and the data corresponding to the coordinate position outside the fluctuation allowable area is the defect determination. The value is less than the threshold value. In the subsequent defect detection processing, whether or not the coordinate position to be determined is included in the fluctuation allowable region is determined using the defect determination threshold D instead of the data of the array H.

  Next, the procedure of defect detection processing will be described with reference to FIG. The processing in steps S400 to S420 and S450 to S490 in FIG. 7 is the same as the processing in steps S200 to S220 and S250 to S290 in FIG.

  In step S430, the defect detection unit 25 refers to G (i) that is the i-th data of the array G. In step S440, the defect detection unit 25 compares G (i) referenced in step S430 with the defect determination threshold D, and determines whether G (i) is greater than or equal to the defect determination threshold D.

  If G (i) referenced in step S430 is greater than or equal to the defect determination threshold D, the point on the m-dimensional space indicated by G (i) is within the fluctuation allowable region. If G (i) referred to in step S430 is less than the defect determination threshold D, the point on the m-dimensional space indicated by G (i) is outside the fluctuation allowable region.

  If the point on the m-dimensional space indicated by G (i) is within the fluctuation allowable region, the process proceeds to step S460. If the point on the m-dimensional space indicated by G (i) is outside the fluctuation allowable region, the process proceeds to step S450.

  As described above, according to the defect detection processing of the present embodiment, it is determined whether or not the coordinate position indicated by the data of the array G is included in the fluctuation allowable region in the m-dimensional space of the time delay coordinate system, and the determination result By detecting the defect based on the above, it is possible to reduce the erroneous detection of the defect even if the dimension of the periodic pattern on the inspection object varies.

  Further, when the reference object and the inspection object are processed by the same manufacturing apparatus or inspection apparatus, the fluctuation pattern of the periodic pattern of the reference object and the inspection object is the same. Therefore, by determining the fluctuation allowable region from the image of the reference object, it is possible to detect defects stably regardless of how the periodic pattern fluctuates during the defect detection process using the image of the inspection object. .

  Note that the fluctuation allowable region determined according to the procedure shown in FIG. 4 or 6 is plotted on the m-dimensional space of the time-delayed coordinate system, and based on the instruction of the user who confirmed the plot on the m-dimensional space. The position and size of the fluctuation allowable area may be changed. Further, the reference object used for determining the fluctuation allowable region may not be a non-defective product, but may be the inspection object itself.

  Even when the fluctuation allowable area is determined from the image of the inspection object itself, the defect allowable area is determined by determining the fluctuation allowable area based on the overall tendency (statistical distribution) of the data of the array H as shown in the modification. It is possible to exclude an area affected by the pixel value of the pixel from the fluctuation allowable area and set an appropriate fluctuation allowable area. Alternatively, based on the instruction of the user who confirmed the plot in the m-dimensional space, an appropriate fluctuation allowable area is set by excluding an area that seems to be affected by the pixel value of the defective pixel from the fluctuation allowable area. Is also possible.

  Next, another method of using the data of the array H indicating the distribution of the fluctuation allowable area will be described. Since the above defect detection processing is performed for each line of data of the image, if the defect detection processing is performed on the entire inspection object, the defect detection processing is performed a plurality of times. Therefore, it is necessary to determine the fluctuation allowable region individually for each defect detection process.

Shall array H is the N generated from the image of one of the reference object, the sequences each _{H 1,} H 2, · · ·, and _{H N.} Control unit 23, based on the data of the sequence H _1, to calculate the barycentric coordinates of the 1 th fluctuation allowable range. Similarly, the control unit 23 calculates the center-of-gravity coordinates of the second, third,... Then, the control unit 23 calculates the average value of the center-of-gravity coordinates of the calculated fluctuation allowable region and stores it in the storage unit 22.

  The fluctuation allowable area obtained from the same reference object is not always used, but the reference object is periodically changed, and a new fluctuation allowable area is determined from the image. The control unit 23 calculates the average value of the center-of-gravity coordinates of the fluctuation allowable region for each reference object, and stores the average value in the storage unit 22.

  The control unit 23 reads the average value of the center-of-gravity coordinates of the fluctuation-permissible areas for a plurality of reference objects from the storage unit 22 periodically during inspection or when receiving an instruction from the user, and corresponds to the average value. Display data for simultaneously (overlapping) displaying the determined position on the m-dimensional space of the time delay coordinate system is generated and output to the display device 3 via the data output unit 26. Based on this display data, the display device 3 displays the distribution of the average value of the center-of-gravity coordinates of the fluctuation allowable region in the m-dimensional space of the time delay coordinate system.

  FIG. 8 is an example in which the distribution of the average value of the center-of-gravity coordinates of the fluctuation allowable region is displayed. A failure of a manufacturing apparatus or an inspection apparatus or its precursor may appear as a change in the fluctuation of the periodic pattern. In FIG. 8, assuming that the center of gravity coordinates of the fluctuation allowable region moves in the direction of the arrow 800 every time, a change with time occurs in the structure of the manufacturing apparatus and the inspection apparatus from the distribution of FIG. It becomes possible to know that. Therefore, it becomes possible to detect a failure at an early stage using the distribution of FIG.

  As described above, the embodiments of the present invention have been described in detail with reference to the drawings. However, the specific configuration is not limited to the above-described embodiments, and includes design changes and the like without departing from the gist of the present invention. . For example, the defect detection method of the present invention can be used in combination with other defect detection methods. As an example, a defect detection process may be executed using the defect detection method of the present invention, and a defect may be detected by a detection method such as another adjacent comparison method at coordinates detected as a defect. As a result, it is possible to detect defects with higher accuracy.

It is a block diagram which shows the structure of the pattern inspection system by one Embodiment of this invention. It is a reference figure for demonstrating the conversion method of the data to the time delay coordinate system in one Embodiment of this invention. It is a reference figure which shows the plot on the m-dimensional space of the time delay coordinate system in one Embodiment of this invention. It is a flowchart which shows the procedure of the determination process of the fluctuation tolerance area | region in one Embodiment of this invention. It is a flowchart which shows the procedure of the defect detection process in one Embodiment of this invention. It is a flowchart which shows the other procedure of the determination process of the fluctuation tolerance area | region in one Embodiment of this invention. It is a flowchart which shows the other procedure of the defect detection process in one Embodiment of this invention. It is a reference figure which shows the mode of a time-dependent change of the gravity center position of the fluctuation tolerance area | region in one Embodiment of this invention. It is a reference figure which shows the luminance change of a periodic pattern. It is a reference figure which shows the luminance difference between the periodic patterns by adjacent comparison.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Imaging device, 2 ... Image processing apparatus, 3 ... Display apparatus, 21 ... Image input part, 22 ... Memory | storage part (memory | storage means), 23 ... Control part (Display data) Generating means), 24 ... time-delayed coordinate system conversion section (data conversion means), 25 ... defect detection section (defect detection means, area determination means), 26 ... data output section

Claims (4)

Data conversion means for converting data composed of pixel values of each pixel constituting an image of the inspection object on which a periodic pattern is formed into data of a time delay coordinate system of dimension m (m is an integer of 2 or more);
a defect detection means for determining whether or not the position indicated by the data of the time delay coordinate system on the m-dimensional space is included in a predetermined area, and detecting a defect based on the determination result;
A defect detection apparatus comprising:   It further comprises area determining means for determining the predetermined area based on the data of the time delay coordinate system of the dimension m obtained by converting data composed of pixel values of each pixel constituting the image of the reference object. The defect detection apparatus according to claim 1. Storage means for storing a plurality of the predetermined areas obtained from the respective images obtained by imaging the reference object at a plurality of different time points;
Display data generating means for generating display data for displaying the distribution in the m-dimensional space of the plurality of predetermined areas stored in the storage means;
The defect detection apparatus according to claim 2, further comprising: Converting data composed of pixel values of each pixel constituting an image of an inspection object on which a periodic pattern is formed into data of a time delay coordinate system of dimension m (m is an integer of 2 or more);
determining whether or not the position indicated by the data of the time delay coordinate system on the m-dimensional space is included in the predetermined area, and detecting a defect based on the determination result;
A defect detection method comprising:

Images