JP2011153832A - Defect review device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a defect review device capable of continuing defect review even if one image processing device is failed, concerning a defect review device for executing defect review or defect classification by using a plurality of image processing devices. <P>SOLUTION: An image processing control part for integrally controlling a plurality of image processing devices is provided, and a survival signal is always transmitted/received between the plurality of image processing devices. When some obstacle is detected in the survival signal (timeout or the like), the image processing control part reassigns image processing devices for executing image processing, relative to image processing executed between the plurality of image processing devices. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a defect review apparatus for identifying a plurality of defects on a wafer used for manufacturing a semiconductor circuit.

  In semiconductor circuits, lowering sales prices, increasing the number of products in small quantities, and shortening delivery time are being promoted. For this reason, in the manufacturing process of a semiconductor circuit, it is required to improve the yield and reduce the manufacturing cost by reducing the chip area by microfabrication. And in order to achieve these, defects, such as a disconnection, a short circuit, and a foreign material adhesion, are discovered at an early stage for every manufacturing process, and the countermeasure is taken.

  Currently, the fine processing of semiconductor circuits is progressing, and the number of elements incorporated in semiconductor circuits is increasing. In addition, the size of wire breaks, short circuits, and foreign objects that cause malfunctions in semiconductor circuits is small, and defects are discovered. The time required for inspection is increasing. Since an increase in inspection time causes an increase in manufacturing cost, a reduction in inspection time is required.

  As an inspection for finding defects, first, for the semiconductor wafer on which the semiconductor circuits are arranged, the defect candidate coordinates of the defect candidate considered as a defect candidate on the semiconductor wafer are obtained by using an appearance inspection device for the entire surface of the wafer. To detect. Next, using an automatic defect review apparatus, the defect candidate coordinates of the semiconductor wafer are enlarged and imaged at a low magnification to obtain a defect candidate image. The defect candidate image is compared with a reference image having no defect to specify the exact defect coordinate of the defect, and the defect is enlarged and imaged at a high magnification based on the accurate defect coordinate to obtain a defect image. Image processing such as image correction may be performed on these captured images. Finally, observation called review is performed on the high-magnification defect image, the cause of the defect is analyzed, and the defect is identified by classification.

  As a conventional technique, a method using a scanning electron microscope has been proposed as a method for acquiring a defect image (see, for example, Patent Document 1). In addition, a defect review apparatus has been proposed that shortens the inspection time by creating a composite reference image that does not include a defect from a low-magnification defect candidate image and reducing the number of times the reference image is captured (see, for example, Patent Document 2). ). Further, a navigation system that stores design data of a semiconductor circuit such as CAD data and sets an imaging / inspection condition including an area to be inspected of a semiconductor wafer based on the design data, and an actual semiconductor according to the imaging / inspection condition. An inspection system provided with a scanning electron microscope that photographs a wafer and performs a length measurement inspection has been proposed (for example, see Patent Document 3 and Patent Document 4).

  Focusing on the automatic defect review in the defect review apparatus, a number of image processes are performed after imaging, such as identification of accurate coordinates of defects, image processing such as image correction, classification of factors causing defects, and the like. In order to shorten the inspection time, it is important to speed up the defect review, that is, speed up the image processing together with the speeding up of the imaging process.

  As a conventional technique, a system has been proposed that includes a plurality of defect detection devices and defect classification devices, and arbitrarily combines them to eliminate a bottleneck in image processing (see, for example, Patent Document 5).

JP 2000-30652 A JP 2007-40910 A JP 2002-328015 A JP 2009-71136 A JP 2006-351636 A

  In Patent Document 5, only a method for configuring a defect classification apparatus and its apparatus are described. Therefore, when processing is concentrated on an image processing device other than the defect classification device, such as a defect detection device or an image correction device, the processing time of the image processing device increases and becomes a bottleneck of inspection time. Since the effect is limited from the viewpoint of speeding up image processing, it becomes a problem to propose a more general solution.

  The present invention pays attention to this problem, and image processing that can avoid degradation in performance of image processing even when the load on specific processing is high for image processing such as defect classification, defect detection, and image correction. A system configuration and a defect review apparatus including the system configuration are provided.

  The present invention is a defect review apparatus that automatically acquires an image by paying attention to a defective portion of a sample, and performs image processing such as defect detection, image correction, and defect classification on an image acquired by the defect review apparatus And a control unit, wherein the image processing control unit controls a plurality of image processing apparatuses having all the image processing functions.

  Further, each image processing apparatus is characterized in that it can execute image processing executed in automatic defect review by an image processing function possessed by each image processing apparatus.

  The overall control unit requests the image processing control unit to perform image processing to be executed during the automatic defect review.

  The image processing control unit can arbitrarily select and instruct the image processing to be executed by each image processing device by distributing the processing to each image processing device for the image processing requested by the overall control unit. It is possible to do.

  According to the present invention, there is provided a processing apparatus configuration capable of improving performance even when specific processing is concentrated, a semiconductor defect review apparatus including the processing apparatus, and an image processing server used from the semiconductor defect review apparatus. Can be provided.

It is a block diagram which shows the basic composition of a SEM type defect review apparatus. It is a figure explaining the hardware constitutions of the image processing apparatus. It is a flowchart for demonstrating an example of the flow of an automatic defect review. FIG. 4A is a diagram conceptually illustrating the image processing function of the image processing apparatus. FIG. 4B is a diagram conceptually showing how the image processing control unit instructs and reports a result to the image processing apparatus and inputs an image. FIG. 5A is a diagram showing an example of the relationship between the defect detection accuracy, which is one of the parameters in defect detection, and the processing time for each number of image processing apparatuses. FIG. 5B is a flowchart illustrating an operation when the image processing control unit assigns a process to the image processing apparatus. FIG. 5C is an example of a table in which the number of image processing apparatuses assigned to image processing and the processing time at that time are obtained. FIGS. 6A and 6B show an example in which the image processing function is assigned to the image processing apparatus by the image processing control unit. It is the figure which showed the method of adding an image processing apparatus dynamically to a defect review apparatus. FIG. 7A shows immediately after the image processing apparatus is added, FIG. 7B shows the recognition status of the image processing apparatus and the assignment status of the image processing function at this time, and FIG. 7C shows the image. Immediately after the survival signal is transmitted from the processing apparatus, FIG. 7D shows a situation in which the survival signal arrives at the image processing control unit and recognizes the image processing apparatus and an image processing function reassignment situation. It is an example of the information which the survival signal which an image processing apparatus transmits. It is the figure which showed the condition which is removing the image processing apparatus dynamically to a defect review apparatus. FIG. 9A shows a situation where the survival signal is interrupted due to the disconnection of the communication path, and FIG. 9B recognizes the image processing apparatus that has lost the survival signal, and performs removal and reassignment of image processing. Shows the situation. It is the flowchart which showed the flow of monitoring of the survival signal in an image processing control part. It is the figure which showed the flow of the process of an image processing control part when an image processing control part detects removal of an image processing apparatus. It is the figure which showed the example of a connection in case a some defect review apparatus exists, and the network between apparatuses. FIG. 12A is a diagram in which a plurality of defect review apparatuses are connected to the network, and FIG. 12B is a diagram in which the defect review apparatus and the image processing server are connected to the network. FIG. 12C is a diagram in which a defect review apparatus and an image processing server that provides an image processing function to the defect review apparatus are connected to a network. FIG. 13A is an example of an information table regarding the use of the image processing apparatus between defect review apparatuses. FIG. 13B is a flowchart showing an operation for enabling an image processing apparatus when using an image processing apparatus provided in another defect review apparatus on the network. FIG. 13C is a flowchart showing an operation when the image processing apparatus is provided to another defect review apparatus on the network.

  FIG. 1 shows a configuration of a semiconductor defect review apparatus according to an embodiment of the present invention. In FIG. 1, a semiconductor defect review apparatus 1 includes an electron gun 101, a lens 102, a deflector 103, an objective lens 104, a sample 105, a stage 106, a secondary particle detector 107, an electron optical system control unit 108, and an A / D conversion. A unit 109, a stage control unit 110, an overall control unit 111, an image processing control unit 112, a display 113, a keyboard 114, a storage device 115, a mouse 116, and the like.

  The electron beam emitted from the electron gun is converged by the lens 102, deflected by the deflector 103, converged by the objective lens 104, and irradiated on the sample 105. From the sample 105, secondary particles 120 such as secondary electrons and reflected electrons are generated according to the shape and material. The generated secondary particles are detected by the secondary particle detector 107, and the signal is converted from an analog signal to a digital signal by the A / D converter 109 to form an SEM image. The formed SEM image is used for image processing such as defect detection and defect classification.

  The lens 102, the deflector 103, and the objective lens 104 are controlled by the electron optical system control unit 108. The sample position control is executed by the stage 106 controlled by the stage control unit 110. The overall control unit 111 interprets inputs from the keyboard 114, the mouse 116, and the storage device 115, and controls the electro-optical system control unit 108, the stage control unit 110, the image processing control unit 112, and the like, and displays as necessary. 113, output the processing contents to the storage device 115. The storage device 115 stores the sent SEM image as image information together with incidental information such as an electro-optical condition when the image is acquired and a recognition number ID of the semiconductor defect review device.

  The image processing control unit 112 is connected to one or more image processing apparatuses 117 via the communication path 118, and instructs the image processing apparatus 117 to receive a command in response to a command from the overall control unit 111. The image processing control unit 112 can instruct image processing execution to any one or more connected image processing apparatuses 117. Image processing such as defect detection, defect classification, and image correction can be performed by any image. It is executed by the processing device 117. The A / D conversion unit 109 equally transfers the acquired SEM image to the image processing apparatus 117 via the image transfer path 119.

  FIG. 2 shows a hardware configuration of the image processing apparatus 117. In FIG. 2A, the image processing apparatus 117 reads an OS (Operating System) stored in the secondary storage medium 204 on the RAM 202 and executes it in accordance with a hardware activation program stored in the ROM 203 after power-on. Next, the image processing program stored in the secondary storage medium 204 is read onto the RAM 202 and executed. This image processing program is a program that provides an image processing function executed during an automatic defect review. When reading the image processing program, only the image processing program that realizes the image processing function required during the automatic defect review is read out and executed on the RAM 202, and if another image processing function is required, the image processing program is executed. Re-reading / execution may be performed. Alternatively, all image processing programs may be read out on the RAM 202 to provide a plurality of image processing functions. The secondary storage medium 204 stores all image processing programs necessary for automatic defect review. The image processing apparatus 117 performs message communication with the outside and image transmission / reception via the network card 205 as necessary. An image capture board 206 may be provided, and an image may be received from the outside via the image capture board 206.

  Note that an OS (Operating System) and an image processing program may not be read from the secondary storage medium 204 but may be read from an accessible storage area on the network, for example, the image processing control unit 112 via the network card 205. . FIG. 2B shows an example of an image processing apparatus 117 that does not have a secondary storage medium, and an image processing function can be provided by a reading method via the network card 205.

  FIG. 3 shows an outline of the flow of semiconductor automatic defect review as an example. This defect detection method is described in Patent Document 1. The overall control unit 111 can refer to defect position information obtained in advance by a defect inspection apparatus or the like (not shown). First, in stage movement S301, based on the defect position information, the overall control unit 111 sends a command to drive the stage to the stage control unit 110, and the stage control unit 110 drives the stage 106 in response to this command. Here, the overall control unit first sends a command to the stage control unit so that the position corresponding to the defect on the chip coordinate system of the chip adjacent to the chip where the defect exists is within the field of view to be imaged at the first magnification. Control the position of the stage. Next, in reference image imaging S302, imaging is performed at the first magnification, and a reference image 311 is obtained.

  Next, in stage movement S303, the stage is driven so that the defect detected by the defect inspection apparatus or the like falls within the field of view to be imaged. Next, defect image 1 (number 312 in the figure) is obtained by imaging at the first magnification in imaging S304 of defect image 1.

  Next, in the defect detection S305, the defect detection is executed by inputting the reference image and the defect image, and the defect position is calculated.

  Next, in imaging S306 of the defect image 2, the imaging magnification is set to a second magnification that is higher than the first magnification. And it images so that a defect position may become a center, and obtains the high-definition defect image 2 (number 313 in a figure). A high-magnification defect image can inspect in detail the shape and surface state of foreign matter and defects.

  Next, using these captured images as input, defect classification is executed in defect classification S307, the type of defect is specified, and the defect image is classified.

  In an example of the flow of the automatic defect review, defect detection and defect classification are executed by the image processing device 117 connected to the image processing control unit 112. Further, image correction can be performed on the captured image by the image processing apparatus 117. Furthermore, image quality can be selected as a parameter in image correction, and detection / classification accuracy can be specified as a parameter for defect detection and defect classification. Further, for defect detection, defect classification, and image correction, it is possible to select whether to execute processing.

  The function of the image processing apparatus 117 will be described with reference to FIG. The image processing apparatus 117 has a plurality of image processing functions necessary for automatic defect review. In the embodiment of FIG. 4, the image processing apparatus 117 has a defect detection function 401, a defect classification function 402, and an image correction function 403, but may further have other image processing functions and related image processing functions. FIG. 4B is a diagram illustrating input / output of information with respect to the image processing apparatus 117. The image processing apparatus 117 receives the command 404 from the image processing control unit 112 that has received a command from the overall control unit 111, and executes image processing. The execution result of the image processing is reported 406 to the image processing control unit 112 through the communication path 118. The image 405 to be processed is acquired from the A / D conversion unit 109 through the image transfer path 119.

  The image processing device 117 periodically transmits a signal to the communication path 118 at regular time intervals (this signal is hereinafter referred to as a survival signal). The image processing control unit 112 and the image processing device 117 can receive a survival signal from another image processing device. Further, the image processing apparatus 117 can transmit the survival signal at an arbitrary timing.

  Next, examples of the present invention will be described more specifically.

(1) Dynamic function assignment to image processing apparatus according to operation setting of automatic defect review The performance required for automatic defect review at that time varies depending on the wafer manufacturing process. For this reason, the defect review apparatus can set various automatic defect review operations such as defect detection, defect classification, presence / absence of image correction, defect detection accuracy, and required image quality. However, if defect detection, defect classification, and image correction are set to be executed with higher accuracy, the time required for the image processing increases and the image processing becomes a bottleneck.

  Therefore, by assigning and executing a plurality of image processing apparatuses for a certain image processing, it is possible to shorten the time of the image processing that becomes a bottleneck. Specifically, the image processing control unit calculates the processing time of each image processing during the automatic defect review based on the setting of the automatic defect review, and executes the image so that the processing time of the entire automatic defect review is minimized. A process is dynamically assigned to each image processing apparatus.

  An example of a method that the overall control unit assigns to image processing is shown in FIG. First, as shown in FIG. 5A, a function for obtaining image processing time for defect review for one defect, or an LUT, is obtained in advance with the parameters (for example, accuracy) of each image processing and the number of image processing apparatuses as variables. Keep it. The defect detection time 511 is an example of the processing time for each image processing apparatus when the defect detection accuracy, which is one of the parameters for automatic defect review, is used as a parameter. As the defect detection accuracy is higher, the defect detection time is longer, and as the number of allocated image processing apparatuses is larger, the defect detection time tends to be shorter. Similarly, in image processing such as defect classification and image correction, when processing with high accuracy and high image quality is performed, the processing time tends to be long, and the processing time tends to be shortened as the number of allocated image processing apparatuses increases. From this automatic defect review parameter setting, the relationship between the number of allocated image processing apparatuses and the processing time of each image processing is uniquely determined.

  FIG. 5B is a flowchart showing the operation when assigning image processing. For image processing assignment, first, the number of currently available image processing apparatuses is acquired (S501). Next, image processing parameters are acquired (S503), and a table of image processing times is created (S504). This operation is repeated for the types of image processing executed in the automatic defect review (S502). Next, the number of image processing apparatuses assigned to each image process is determined so that the total image processing time is minimized (S505). In accordance with this determination, image processing is assigned to the image processing apparatus (S506).

  FIG. 5C is an example of a table that is created in S504 and that determines the number of image processing apparatuses allocated to image processing and the processing time at that time. For example, referring to the defect detection time table 512, it can be seen that when three image processing apparatuses are assigned to defect detection, 80 processing times are required. Therefore, using the defect detection time table 512, the defect classification table 513, and the image correction table 514, the combination of the number of image processing apparatuses to be allocated is set to S505 so that the time for executing all image processing is minimized. To decide. In the case of FIG. 5C, the shortest processing time is the case where one unit is assigned for defect detection, three units are assigned for defect classification, and one unit is assigned for image correction. As a method for calculating this combination, for example, for all the combinations, a combination that has the shortest processing time may be searched, or a combination may be efficiently obtained using dynamic programming or the like.

  In the example of FIG. 6A, only one of the image processing functions of the image processing apparatus 117 is operating, but one unit is provided for the defect detection function 401, three units are provided for the defect classification function 402, and an image correction function 403 is provided. One image processing apparatus 117 is assigned to each of the two, and constitutes an image processing system of the defect review apparatus. When specific image processing is not executed, the number of image processing devices 117 assigned to the image processing may be zero. In the case of FIG. 6B, the number of image processing apparatuses to which the defect classification function 402 is assigned is zero.

(2) Dynamic Addition / Removal of Image Processing Device in Defect Reviewing Device A method for dynamically adding an image processing device to the defect reviewing device will be described with reference to FIG. The dynamic addition means that a new image processing device is attached to the active defect review device so that the image processing control unit 112 can assign image processing during automatic defect review. However, this does not limit the dynamic addition meaning to physical installation of the image processing device hardware. Further examples will become apparent later.

  FIG. 7A shows a state immediately after the image processing apparatus 701 is added to the defect review apparatus. The communication path 118 and the image transfer path 119 are appropriately connected to the image processing apparatus 701 when added. FIG. 7B is a diagram showing a list of image processing devices to which the image processing control unit 112 grasps and assigns image processing immediately after the image processing device 701 is added. Immediately after the addition of the image processing device 701, the image processing control unit 112 has not detected the presence of the image processing device 701, and the image processing function is not assigned. FIG. 7C is a diagram showing a situation immediately after the image processing apparatus 701 transmits a survival signal to the communication path 118, which is a feature of the present invention. The signal transmitted by the image processing device 701 reaches the image processing control unit 112 and each image processing device 117, and receives a survival signal of the image processing device 701. Based on this survival signal, the image processing control unit 112 detects that the image processing apparatus 701 has been added to the defect review apparatus. FIG. 7D is a diagram illustrating a state of image processing function assignment when the image processing control unit 112 executes automatic defect review after the addition detection of the image processing apparatus 701. The image processing apparatus 701 is recognized, and an image processing function in automatic defect review is assigned.

  The survival signal transmitted to the communication path 118 by the image processing apparatus includes information for distinguishing the signal transmission source. FIG. 8 shows a structural example of the survival signal 801. As information of the transmission source, for example, there is information such as the IP address 802 of the transmission source, a unique image processing device number 803 assigned in advance to the image processing device, transmission time 804, device status 805 including current operation information, and the like. The overall control unit refers to the information specific to the image processing device included in the survival signal, and if it is not included in the image processing device recognized at that time, the image processing device is newly added. recognize.

  Further, a method for dynamically removing the image processing apparatus from the defect review apparatus will be described. FIG. 9A is a diagram illustrating a state in which the communication path 118 connected to the image processing apparatus 901 is physically or logically disconnected. At this time, the image processing apparatus 901 cannot transmit the survival signal 801 to the communication path 118. On the other hand, since the image processing control unit 112 does not receive the survival signal 801 for a certain period of time, a timeout of arrival of the survival signal for the image processing apparatus 901 occurs. From this timeout, it is recognized that the image processing apparatus 901 has been removed, and image processing is not assigned by automatic defect review (FIG. 9B).

  Even when the survival signal 801 arrives at the image processing control unit 112, if the apparatus status 805 includes information indicating that the image processing cannot be performed or the image processing allocation is rejected, the survival signal is transmitted. The original image processing apparatus treats it as removed and does not assign image processing. However, if the survival signal from the image processing apparatus arrives again and the apparatus status 805 does not include information indicating that image processing cannot be performed or image processing allocation is rejected, the image processing apparatus Treated as added to.

  FIG. 10 is a flowchart showing a flow of monitoring a survival signal in the image processing control unit 112. In this flow, the image processing control unit 112 recognizes that the connection status of the image processing apparatus has changed. First, the image processing control unit 112 monitors an arrival of a survival signal or a time-out event (S1001). If any event occurs, it is determined what the generated event is (S1002), and if it is timed out, the corresponding image processing apparatus is dynamically removed (S1007). In the case of the survival signal arrival, it is confirmed whether the source of the arrived survival signal is currently recognized by the image processing control unit 112 (S1003), and if it is not recognized, it is recognized as a new image processing apparatus. And add to the event monitoring. If the transmission source of the arrived survival signal is recognized, the survival signal arrival timer is reset (S1004), and the time until timeout is extended. Next, information such as refusal of image processing assignment or inability to execute among the information included in the survival signal is referred to, and it is confirmed whether image processing assignment is possible for the image processing apparatus (S1005). If assignment is possible, the process returns to event monitoring (S1001) again. If image processing cannot be assigned, the corresponding image processing apparatus is dynamically removed (S1007), and the process returns to event monitoring (S1001).

  The reasons for the dynamic removal of the image processing apparatus are (I) physical removal of the image processing apparatus, (II) hardware or software failure of the image processing apparatus, and (III) image on the image processing apparatus side. For example, rejection of process assignment. Of these, when (I) and (II) are the causes, there is no connection between the communication path 118 and the image processing apparatus 901, or normal communication is not possible. When such a state occurs during image processing of the image processing apparatus 901, the image processing control unit 112 reports that the image processing result from the image processing apparatus 901 is an invalid result or is not reported, and image processing control is performed. The image processing result that should be received by the unit 112 is lost. On the other hand, in the case of (III), the connection between the communication path 118 and the image processing apparatus 901 is normal, and the assignment of the image processing function is simply not permitted. Therefore, it can be expected that the image processing result is correctly reported.

  FIG. 11 shows an operation when dynamic removal of the image processing apparatus is detected. When the dynamic removal of the image processing apparatus is detected, the image processing control unit first deletes the removed image processing apparatus from the recognition list of the image processing apparatus (S1101). Next, it is determined whether an automatic defect review is currently being performed (S1102). If not currently in automatic defect review, no further processing is required for dynamic removal. However, if the automatic defect review is being performed, it is determined whether the image processing apparatus is performing image processing (S1103). If image processing is not in progress, the image processing control unit immediately reassigns image processing to all image processing apparatuses (S1106). This is because the optimum allocation changes due to the change in the number of image processing apparatuses. However, if the image is being processed, a survival signal is confirmed (S1104). When the apparatus state of the survival signal is information indicating that image processing allocation is rejected, the image processing result being executed can be expected to be reported correctly, so the image processing control unit immediately performs image processing for all image processing apparatuses. Are reassigned (S1106). On the other hand, if the apparatus status of the survival signal is information indicating that image processing cannot be performed, or if the survival signal does not arrive, it is determined that the image processing result being executed has been lost, and the automatic defect review is abnormally stopped (S1105). .

(3) Use of Image Processing Apparatus Provided by Other Defect Review Apparatus and Image Processing Server A function for executing automatic defect review using an image processing apparatus provided by another defect review apparatus will be described. FIG. 10 is a diagram showing a connection example and a network between devices when there are a plurality of defect review devices. FIG. 10A shows a defect review device 1202, 1203, 1204 connected to the network 1201. Each defect review apparatus includes one or more image processing apparatuses 117, and the defect review apparatuses can report current apparatus information to each other via the network 1201. This apparatus information includes information on an image processing apparatus that can be provided by the defect review apparatus. The communication between the defect review apparatuses can be realized by, for example, one-to-one communication with a defect review apparatus registered in advance or N-to-one communication in which a signal is transmitted to a plurality of defect review apparatuses. it can. Accordingly, when the defect review apparatus according to the present invention performs dynamic function assignment of image processing, the number of image processing apparatuses that can be provided by other defect review apparatuses and the performance of the image processing apparatus are determined between the defect review apparatuses. A defect review apparatus that communicates and wants to use an image processing apparatus for automatic defect review or the like can use the image processing apparatus of the defect review apparatus. FIG. 13A is an example of an information table regarding the use of the image processing apparatus between defect review apparatuses. There are the number of currently available image processing obtained by communication between defect review apparatuses, processing delay information for taking into account overhead generated during processing, and the like.

  In FIG. 12B, an example in which the defect review apparatus 1202 uses the image processing apparatus 1205 in the defect review apparatus 1203 and assigns an image processing function is shown in FIGS. 13B and 13C as a flowchart. FIG. 13B is a flowchart on the defect review apparatus 1202 side, and FIG. 13C is a flowchart on the defect review apparatus 1203 side. The defect review apparatus 1203 has transmitted apparatus information to the defect review apparatus 1202 in advance. Hereinafter, the defect review device 1202 is described as a use side, and the defect review device 1203 is described as a providing side.

  First, the use side 1202 inquires of the providing side 1203 about the use permission of the image processing apparatus (S1301), and determines whether or not the image processing apparatus can be used from the response to the inquiry (S1302). When the use is not permitted, the image processing is not assigned by the use side 1202. When the use is permitted, the communication path in the use side 1202 and the image processing apparatus 1205 are logically connected via the network 1201 (S1303). With this process, the image processing apparatus 1205 is logically removed from the providing side 1203 and is logically connected to the using side 1202. Next, the image processing control unit provided on the use side 1202 dynamically adds the image processing apparatus 1205 (S1304). However, the use of the image processing device provided in the other defect review device has an overhead such as image transfer over the network and communication delay, so compared to the case of using the image processing device inside the defect review device, Performance decreases. Therefore, when the use side 1202 assigns an image processing function to the image processing apparatus 1205 provided on the providing side 1203, the image processing function assignment is calculated in consideration of the processing delay due to overhead.

  On the other hand, the providing side 1203 receives a use permission inquiry from the use side 1202 (S1305), and determines whether to allow use from outside the defect review apparatus (S1306). If it does not permit, it responds as disapproval (S1308). In the case of permission, a permission response is made (S1307), and the image processing apparatus to be used is removed from the internal network (S1309). Next, the image processing apparatus to be used is logically connected to the internal network of the use side 1202 (S1310), and the process ends.

  When canceling the use of the image processing apparatus 1205 from the use side 1202, the use side 1202 notifies the image processing apparatus 1205 of the use cancellation directly through a logical communication path via the network 1201 or the providing side 1203 as a proxy. To do. Thereafter, the use side 1202 releases the logical connection with the image processing apparatus 1205 and performs dynamic removal. Further, the image processing device 1205 is dynamically added to the providing side 1203.

  When the image processing apparatus 1205 side cancels the use from the use side 1202, the image processing apparatus 1205 releases the logical connection with the use side 1202 and dynamically adds to the providing side 1203.

  Further, as shown in FIG. 12B, image processing is also performed by using image processing servers 1206 and 1207 each having a defect detection function 401, a defect classification function 402, and an image correction function 403, which are image processes performed during automatic defect review. Can also be done. The image processing servers 1206 and 1207 are connected on the network 1201. The image processing servers 1206 and 1207 provide only an image processing function to the defect review apparatus, and do not include an apparatus for imaging a sample. The image processing server may include a plurality of image processing apparatuses 117 to provide an image processing function, or may include an image processing system 1209 that is capable of providing the image processing function but is not the image processing apparatus 117. The image processing server can report the current apparatus information including the operation status of the image processing apparatus to the defect review apparatus via the network 1201.

  In the present invention, the defect review apparatus uses the image processing apparatus 117 of the image processing server, so that the image processing server 1207 does not have to be provided like the defect review apparatus 1208. In that case, the image processing function is the image processing server 1206. , 1207 or both. Alternatively, an image processing device 117 provided in another defect review device may be used, and an image processing server may be used at the same time. In addition, an image processing server on the network 1201 can be used even by a defect review apparatus including the image processing apparatus 117 such as the defect review apparatuses 1202, 1203, and 1204.

  With this function, the following effects can be obtained.

Optimize image processing of defect review apparatus The present invention assigns an image processing function to each image processing apparatus so that the defect review apparatus minimizes the automatic defect review time. The design burden regarding the performance of the image processing function is reduced, and the maximum performance can be obtained.

The image manufacturer can easily expand the image processing performance of the defect review apparatus. The apparatus manufacturer can add an image processing apparatus simply by connecting the communication path 118 and the image transfer path 119. It is possible to easily improve the performance. Therefore, flexible customization is possible according to the performance and cost required by the device user. Furthermore, when a plurality of defect review apparatuses are connected via a network, an image processing apparatus included in another defect review apparatus can be used.

-Obtaining redundancy of image processing devices of defect review devices Even when some image processing devices fail, image processing functions are automatically assigned to image processing devices that can be used at the start of automatic defect review. Although the performance degradation of the image processing is recognized, the defect review apparatus can continue to operate.

1 Semiconductor defect review device 1
Reference Signs List 101 electron gun 102 lens 103 deflector 104 objective lens 105 sample 106 stage 107 secondary particle detector 108 electron optical system control unit 109 A / D conversion unit 110 stage control unit 111 overall control unit 112 image processing control unit 113 display 114 keyboard 115 Storage Device 116 Mouse 117, 701, 901, 1205 Image Processing Device 119 Image Transfer Path 120 Secondary Particle 202 RAM
203 ROM
204 Secondary storage medium 205 Network card 206 Image capture board 311 Reference image 312 Defective image 1
313 Defect image 2
401 Defect detection function 402 Defect classification function 403 Image correction function 405 Processing target image 511 Defect detection time 512 Defect detection time table 513 Defect classification table 514 Image correction table 801 Survival signal 802 IP address 803 Image processing apparatus number 804 Transmission time 805 Device status 1201 Network 1202, 1203, 1204, 1208 Defect review device 1206, 1207 Image processing server 1209 Image processing system

Claims (6)

In a defect review device that automatically acquires an image focusing on the defective part of the sample,
An imaging means for acquiring an image of the defective portion and outputting it as image information;
A plurality of image processing means for processing the image information;
A plurality of image processing means for performing a plurality of image processing such as defect detection, image correction, defect classification, etc. on the image acquired by the defect review device;
A defect review apparatus comprising: an image processing control unit that controls the plurality of image processing means. The defect review apparatus according to claim 1,
The defect review apparatus, wherein the image processing apparatus transmits a signal including information capable of identifying an individual such as an operation state of the apparatus and its own apparatus number periodically at an arbitrary timing or at a constant time interval. The defect review apparatus according to claim 1,
The overall control unit requests the image processing control unit to perform image processing to be executed during an automatic defect review. The defect review apparatus according to claim 1,
The image processing control unit arbitrarily selects and instructs image processing to be executed by each image processing device by distributing the processing to each image processing device for the image processing requested by the overall control unit. Defect review equipment. The defect review apparatus according to claim 1,
A defect review apparatus characterized in that an apparatus for performing image processing can be added without stopping operation even when the defect review apparatus is in operation.   6. The defect review apparatus according to claim 5, wherein a part of the image processing apparatus can be removed in place of the addition.

Images