CN116430679A - Hot spot detection method and device - Google Patents

Abstract

The application relates to the field of hot spot identification, in particular to a hot spot detection method and a hot spot detection device, which can solve the problem of poor accuracy in hot spot detection in a photoetching simulation mode to a certain extent. The hot spot detection method comprises the following steps: storing a plurality of first images including circuit layouts into an image library, wherein the circuit layouts contain hot spots; determining a matching image matched with an image to be detected in the image library, wherein the first image comprises the matching image, and the image to be detected comprises a wafer needing hot spot detection; acquiring a feature vector of the matched image; and inputting the feature vector into a preset detection model, and determining the type and the position of the hot spot in the wafer based on the output of the detection model.

Description

Hot spot detection method and device

Technical Field

The present application relates to the field of hotspot identification, and in particular, to a hotspot detection method and apparatus.

Background

One of the essential links in the chip manufacturing process is to carry out photoetching on a wafer through a photoetching machine, namely, the photoetching machine removes specific parts on the surface of the wafer through a series of process steps according to a designed circuit layout, the graph on the surface of the wafer is a graph shown by the circuit layout, and the wafer after photoetching forms a final chip through subsequent processing. At present, with the continuous improvement of the requirements on the performance and the integration level of chips, the wavelength of photolithography performed by a photoetching machine is continuously reduced, so that the optical proximity effect (Optical Proximity Effect, OPE) caused by the wavelength is capable of causing the distortion of the pattern on the surface of a wafer, and the surface of the wafer after photoetching has defects such as bridging (bridging) and breaking (breaking), which are called as photoetching hot spots for short. The existence of the hot spot causes errors of the chip and affects the normal use of the chip, so that the hot spot existing after the wafer photoetching is necessary to be detected so as to process the wafer with problems in time.

The current hot spot detection is generally performed based on lithography simulation, specifically, a series of initial process parameters are set, the lithography process of a wafer is simulated, in the process, a series of initial process parameters are fitted for multiple times, the process parameters are corrected continuously according to the fitting result until a final simulation result is obtained, the simulation result is a wafer virtual pattern obtained after the lithography process is simulated, the simulation result is compared with a real pattern of the wafer object surface obtained by carrying out lithography on the wafer according to the same circuit layout, the distinguishing points of the two are determined, and the position of the distinguishing points corresponding to the wafer object is the position of the hot spot.

However, the method for performing hot spot detection through lithography simulation needs to perform fitting correction on a large number of process parameters for multiple times to obtain a simulation result, and the parameters in the process are usually related to each other, so that the continuous fitting correction process is very prone to error, the simulation result is inaccurate, and the accuracy of the hot spot position in the finally determined wafer real object is poor.

Disclosure of Invention

In order to solve the problem of poor accuracy in hot spot detection in a photoetching simulation mode, the application provides a hot spot detection method and device.

A first aspect of the present application provides a hotspot detection method, the method comprising:

storing a plurality of first images comprising circuit layouts into an image library, wherein the circuit layouts comprise hot spots, and the hot spots in at least two first images are different;

determining a matching image matched with an image to be detected in the image library, wherein the first image comprises the matching image, and the image to be detected comprises a wafer needing hot spot detection;

acquiring a feature vector of the matched image;

and inputting the feature vector into a preset detection model, and determining the type and the position of the hot spot in the wafer based on the output of the detection model.

In some possible implementations, the detection model includes at least one classifier, and after storing the plurality of first images including the circuit layout in the image library, the method further includes:

dividing the first image into at least one image set, one of the image sets corresponding to one of the classifiers;

the inputting the feature vector into a preset detection model comprises the following steps:

and inputting the feature vector into the classifier corresponding to the image set where the matched image is located.

In some possible implementations, the dividing the first image into at least one image set includes:

and dividing the first image according to the type of the hot spot and/or based on the area around the hot spot in the circuit layout.

In some possible implementations, before the dividing the first image according to the type of the hotspot, the method further includes:

determining a hotspot type in the first image based on a difference between the area of the circuit layout and the actual area of the circuit layout in the first image;

if the difference value between the area of the circuit layout and the actual area is a first preset value, and the area of the circuit layout is smaller than the actual area, the hot spot type is a wire breakage type;

if the difference value between the area of the circuit layout and the actual area is a second preset value, and the area of the circuit layout is larger than the actual area, the hot spot type is a line bridging type;

and if the difference value between the area of the circuit layout and the actual area is a third preset value, the hot spot type is a contact hole type, wherein the third preset value is smaller than the second preset value, and the third preset value is smaller than the first preset value.

In some possible implementations, the dividing the first image based on the area around the hot spot in the circuit layout includes:

determining a first area by intercepting areas around a hot spot position in the first image;

and determining that the first images with the same first area are the same type of image set.

In some possible implementations, the feature vector includes at least any two of the following parameters:

the shortest distance from the center of the hot spot to the edge of the matching image, the straight line shortest distance from the center of the hot spot to the edge of the circuit layout, the longest distance from the center of the hot spot to the edge of the circuit layout, the shortest distance from the edge of the matching image to the edge of the circuit layout, the longest distance from the edge of the matching image to the edge of the circuit layout and the shortest distance from the center of the hot spot to the 4 corners of the circuit layout.

In some possible implementations, the obtaining the matching image in the image library, which matches the image to be tested, includes:

determining the matching score of the circuit layout graph in the first image and the circuit layout graph in the image to be detected;

And if the matching score is greater than a first threshold value, determining the first image as the matching image.

In some possible implementations, the method further includes:

if no matching image matched with the image to be detected exists in the image library, inputting the feature vector of the image to be detected into a preset detection model, and obtaining a detection image output by the detection model;

and storing the detection image into the image library.

In some possible implementations, the training process of the detection model includes:

acquiring a plurality of sample images and feature vectors corresponding to the sample images, wherein the sample images comprise real hot spot positions and real hot spot types;

sequentially inputting the feature vectors of the sample images into a detection model to be trained, and predicting the sample hot spot positions and the sample hot spot types in the sample images through the detection model to be trained;

and carrying out iterative training on the detection model to be trained according to the real hot spot position, the sample hot spot position, the real hot spot type and the sample hot spot type, and obtaining the detection model after training is finished.

A second aspect of the present application provides a hotspot detection apparatus, the apparatus comprising:

The image storage module is used for storing a plurality of first images comprising circuit layouts into an image library, wherein the circuit layouts comprise hot spots;

the image matching module is used for determining a matching image matched with an image to be detected in the image library, wherein the first image comprises the matching image, and the image to be detected comprises a wafer needing hot spot detection;

the feature vector acquisition module is used for acquiring the feature vector of the matched image;

and a detection module: and the feature vector is used for inputting the feature vector into a preset detection model, and the type and the position of the hot spot in the wafer are determined based on the output of the detection model.

A third aspect of the present application provides a hotspot detection system, the system comprising:

a memory storing a computer program, and a processor implementing the steps of the method of any one of the first aspects above when the computer program is called from the memory and executed by the processor.

A fourth aspect of the present application provides a computer readable storage medium having stored therein at least one computer program which when executed by a processor performs the steps of the method of any of the first aspects above.

The technical scheme provided by the application can at least achieve the following beneficial effects:

the method comprises the steps of storing a plurality of first images into an image library so as to find a matched image matched with an image to be detected in the image library, inputting the feature vector of the matched image into a preset detection model, detecting the feature vector through the preset detection model, and determining the position of a hot spot in the detected image.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief description will be given below of the drawings that are needed in the embodiments or the prior art descriptions, it being obvious that the drawings in the following description are some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort to a person skilled in the art.

FIG. 1 is a schematic diagram of a computer device according to an exemplary embodiment of the present application;

FIG. 2 is a flow chart of a method for hot spot detection according to an exemplary embodiment of the present application;

FIG. 3 is a flowchart illustrating a hot spot detection method according to an exemplary embodiment of the present application;

FIG. 4 is a schematic diagram of a hotspot shown in an exemplary embodiment of the present application;

FIG. 5 is a schematic diagram of a hotspot shown in an exemplary embodiment of the present application;

FIG. 6 is a schematic diagram of a hotspot shown in an exemplary embodiment of the present application;

FIG. 7 is a flowchart illustrating a first image dividing step in a hot spot detection method according to an exemplary embodiment of the present application;

FIG. 8A is a first image schematic diagram including a first region, as shown in an exemplary embodiment of the present application;

FIG. 8B is a first image schematic diagram including a first region, as shown in an exemplary embodiment of the present application;

FIG. 8C is a first image schematic diagram including a first region, as shown in an exemplary embodiment of the present application;

FIG. 8D is a first image schematic diagram including a first region, as shown in an exemplary embodiment of the present application;

FIG. 9 is a schematic diagram of parameters in a feature vector according to an exemplary embodiment of the present application;

FIG. 10 is a flowchart illustrating a test model training process in a hot spot test method according to an exemplary embodiment of the present application;

FIG. 11 is a schematic diagram illustrating a hot spot detection method detection overall process according to an exemplary embodiment of the present application;

fig. 12 is a schematic structural diagram of a hot spot detection device according to an exemplary embodiment of the present application.

Detailed Description

For purposes of making the objects, embodiments and advantages of the present application more apparent, the exemplary embodiments of the present application will be described in detail and fully in connection with the accompanying drawings in which exemplary embodiments of the present application are shown, it being understood that the exemplary embodiments described are only some, but not all, of the examples of the present application, and it is to be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the application.

It should be noted that the brief description of the terms in the present application is only for convenience in understanding the embodiments described below, and is not intended to limit the embodiments of the present application. Unless otherwise indicated, these terms should be construed in their ordinary and customary meaning.

The terms "first," second, "" third and the like in the description and in the claims and in the above-described figures are used for distinguishing between similar or similar objects or entities and not necessarily for limiting a particular order or sequence, unless otherwise indicated. It is to be understood that the terms so used are interchangeable under appropriate circumstances.

The terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a product or apparatus that comprises a list of elements is not necessarily limited to all elements explicitly listed, but may include other elements not expressly listed or inherent to such product or apparatus.

In order to clearly describe the technical solutions of the embodiments of the present application, some concepts related to the present application will be described below first.

(1) Wafer (Wafer)

Wafer refers to a silicon wafer used for manufacturing silicon semiconductor circuits, the original material of which is silicon. The high-purity polycrystalline silicon is dissolved and then doped with silicon crystal seed, and then slowly pulled out to form cylindrical monocrystalline silicon. After grinding, polishing and slicing, the silicon wafer, i.e. the wafer, is formed, and the wafer is generally subjected to photolithography processing in the process of processing.

(2) Photolithographic processing

The photoetching process is a core step in the manufacturing process of an integrated circuit, and a specific part of the surface of a silicon wafer is removed through a series of process steps, so that the transfer of a design layout pattern to each layer of material on the surface of the silicon wafer is realized, and a laminated structure comprising devices is formed on the wafer. A typical lithographic processing system is usually composed of four basic elements, namely a light source system, a mask plate, a projection system and a wafer, wherein the mask plate is a pattern template used in the lithographic process and carries thereon designed circuit pattern, i.e. the mask plate solidifies the data of each layer in the circuit pattern. In a specific photoetching process, light waves emitted by a light source system pass through a transparent area of a mask plate and then carry modulation information of a circuit layout, areas with different exposure intensities are formed on the surface of a wafer covered with photoresist, and at the moment, a photoetching imaging pattern is formed on the surface of the wafer and is consistent with the pattern of the circuit layout.

It should be noted that, in consideration of the process error, the patterns described herein are consistent, and it is understood that the overlap ratio of the photolithography pattern and the circuit layout pattern is within a preset error range, and it is not required that both are identical.

(3) Optical proximity effect (Optical Proximity Effect, OPE)

With the development of semiconductor technology, the feature size of a semiconductor device is smaller and smaller, and in the photolithography process, when the feature size of the semiconductor device is close to or even smaller than the wavelength of a light wave used in the photolithography process, due to diffraction and interference phenomena of light, a certain deformation exists between a photolithography pattern obtained on an actual wafer and a pattern on a mask plate, for example, the photolithography pattern on the surface of the wafer has distortion conditions such as line end shortening, line width unevenness or corner rounding, and the phenomenon is called optical proximity effect (Optical Proximity Effect, OPE for short).

(4) Hot spot

The region of the wafer, which is subject to defects due to the distortion of the pattern of the lithography image caused by the optical proximity effect, is called a lithography hotspot, which is called a hotspot for short, and the hotspot has a great adverse effect on the manufacturing yield of the subsequent integrated circuit.

(5) Neural Networks (ANN)

Neural networks are a subset of machine learning and are the core of deep learning algorithms. The name and structure of the biological neuron are inspired by the human brain, and imitate the mutual transmission mode of biological neuron signals.

The neural network consists of a layer of nodes comprising an input layer, one or more hidden layers and an output layer, each node also called an artificial neuron, connected to another node, with associated weights and thresholds. If the output of any single node is above a specified threshold, that node will be activated and send the data to the next layer of the network. Otherwise, the data is not passed on to the next layer of the network.

Next, application scenarios and implementation environments of the embodiments of the present application are described.

In the process of transferring the pattern of the circuit layout onto the wafer, the pattern photoetched onto the wafer generates larger distortion due to the influence of the optical proximity effect, so in order to improve the phenomenon, a phase shift mask, optical proximity correction or other modes are generally adopted to optimize the pattern structure photoetched on the wafer. However, in the sub-wavelength lithography, the problem of hot spot on the wafer is still unavoidable, and the wafer seriously affects the yield of the subsequent integrated circuit, so that it is necessary to perform hot spot detection on the wafer. The hot spot detection may be defined as precisely locating hot spots existing in the layout within an acceptable time range.

At present, the main mode of hot spot detection is to carry out hot spot detection based on lithography simulation, specifically, a series of initial process parameters are set for the same lithography model as the wafer to be subjected to lithography, wherein the initial process parameters are specific lithography process parameters such as exposure intensity, focal depth and the like; and then simulating the photoetching model to enable the photoetching model to simulate the photoetching process of the wafer under the same photoetching process condition. In the simulation process, a series of initial process parameters are fitted for multiple times, and the process parameters are continuously corrected according to the fitting result until a final simulation result is obtained, wherein the simulation result is a wafer virtual pattern obtained after the photoetching process is simulated. Comparing the simulation result with the real pattern of the wafer object surface obtained by photoetching the wafer according to the same circuit layout, and determining the distinguishing point of the simulation result and the real pattern, wherein the position of the distinguishing point corresponding to the wafer object is the position of the hot spot.

In the process of determining the distinguishing points, hot spot areas possibly having problems in the wafer real object can be screened out according to parameters such as errors of edge positions between the virtual patterns and the real patterns.

However, in the above process, the method for performing hot spot detection by photolithography simulation needs to perform fitting correction on a large number of process parameters for multiple times to obtain a simulation result, and the parameters in the process are usually related to each other, so that the continuous fitting correction process is prone to error, and the simulation result is inaccurate, so that the accuracy of the hot spot position in the finally determined wafer real object is poor.

Therefore, based on the problem of poor accuracy of the detected hot spot position, the application provides a hot spot detection method, by determining an image library, conveniently acquiring a matched image of an image to be detected from the image library, inputting a feature vector of the matched image into a preset detection model, and acquiring a hot spot type and a position of the hot spot output by the preset detection model, in the process, the feature vector of the matched image is input into the preset detection model for detection, the preset detection model does not need to extract features in each image to be detected, so that the data processing amount is reduced, and the final detection result is more accurate.

The hot spot detection method can utilize computer equipment to store a trained detection model, and determine the type and the position of the hot spot in the wafer based on the feature vector of the matched image by running the detection model.

In some embodiments, the architecture of the computer device is as shown in FIG. 1, the computer device 100 comprising at least one processor 110, a memory 120, a communication bus 130, and at least one communication interface 140.

The processor 110 may be a general purpose CPU, network processor (Network Processor, NP), microprocessor, or may be one or more integrated circuits for implementing aspects of the present Application, such as Application-specific integrated circuits (ASIC), programmable logic devices (Programmable Logic Device, PLD), or a combination thereof.

The PLD may be a complex programmable logic device (Complex Programmable Logic Device, CPLD), a Field programmable gate array (Field-Programmable Gate Array, FPGA), general array logic (Generic Array Logic, GAL), or any combination thereof.

Wherein the computer device 100 may include a plurality of processors 110, the processors 110 may include one or more CPUs. Each of the processors 110 may be a single-core processor (single-CPU) or a multi-core processor (multi-CPU).

It is noted that the processor 110 herein may refer to one or more devices, circuits, and/or processing cores for processing data (e.g., computer program instructions).

Memory 120 may be a Read-Only Memory (ROM) or other type of static storage device that may store static information and instructions; random access memory (Random Access Memory, RAM) or other types of dynamic storage devices that can store information and instructions; but not limited to, an electrically erasable programmable read-Only Memory (EEPROM), a compact disk read-Only Memory (CD-ROM) or other optical disk storage, optical disk storage (including compact disk, laser disk, optical disk, digital versatile disk, blu-ray disk, etc.), magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.

It should be noted that, the memory 120 may be independent and connected to the processor 110 through the communication bus 130. Of course, the memory 120 may also be integrated with the processor 110.

Communication bus 130 is used to transfer information between components (e.g., between a processor and a memory), and communication bus 130 may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, fig. 1 is illustrated with only one communication bus, but not with only one bus or one type of bus.

The communication interface 140 is used for the computer device 100 to communicate with other devices or communication networks. The communication interface 140 includes a wired communication interface or a wireless communication interface. The wired communication interface may be, for example, an ethernet interface. The ethernet interface may be an optical interface, an electrical interface, or a combination thereof. The wireless communication interface may be a wireless local area network (Wireless Local Area Networks, WLAN) interface, a cellular network communication interface, a combination thereof, or the like.

In some embodiments, computer device 100 may also include output devices and input devices (not shown in FIG. 1). Wherein the output device is in communication with the processor 110 to display information in a variety of ways, for example, the output device may be a liquid crystal display (Liquid Crystal Display, LCD), a light emitting diode (Light Emitting Diode, LED) display device, a Cathode Ray Tube (CRT) display device, or a projector (projector), etc.; the input device, which may be, for example, a mouse, a keyboard, a touch screen device, a sensing device, or the like, communicates with the processor 110 to receive user input in a variety of ways.

In some embodiments, the memory 120 may be used to store a computer program that performs aspects of the present application, and the processor 110 may execute the computer program stored in the memory 120. For example, the computer device 100 may invoke and execute a computer program stored in the memory 120 by the processor 110 to implement the steps of the optical proximity correction method provided by embodiments of the present application.

It should be understood that the hot spot detection method provided in the present application may also be applied to a hot spot detection apparatus, where the hot spot detection correction apparatus may be implemented as part or all of a processor by using software, hardware, or a combination of software and hardware, so as to be integrated in any computer device.

Next, the technical solutions of the present application and how the technical solutions of the present application solve the above technical problems will be specifically described by way of examples with reference to the accompanying drawings. Embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. It will be apparent that the described embodiments are some, but not all, of the embodiments of the present application.

Referring to fig. 2, the present application provides a hot spot detection method, which is applied to the computer device 100 shown in fig. 1 and illustrated by way of example, and the method may include the following steps:

Step 210: storing a plurality of first images including circuit layouts into an image library, wherein the circuit layouts include hot spots, and the hot spots in at least two first images are different.

The conventional image acquisition equipment can be used for acquiring images of a plurality of circuit layouts containing hot spots to acquire a plurality of first images.

The circuit layout included in the first image in the application can be a complete circuit layout of a single-layer process of an integrated circuit device, or can be a circuit layout of each unit circuit in the circuit layout, and the circuit layouts in each first image can be the same or different, and the embodiment of the application does not limit the circuit layout.

It should be noted that, the types of the hot spots in the at least two first images are different or the positions of the hot spots are different, or the types and the positions of the hot spots are different.

In some possible embodiments, the same picture preprocessing may be performed on each first image, such as performing an image denoising operation on each first image, so that the first images can be clearer.

Step 220, determining a matching image matching with an image to be detected in the image library, wherein the first image comprises the matching image, and the image to be detected comprises a wafer to be subjected to hot spot detection.

The image to be detected can be obtained by shooting a wafer needing hot spot detection through a conventional image acquisition device, and the image to be detected can be subjected to the same picture preprocessing operation as the first image. In the process of determining the matching image, the method can be performed by an identification algorithm for identifying the region in the image to be detected and a search algorithm for searching the matching image in the image library, and the matching image acquisition process has good accuracy.

Specifically, in some possible embodiments, as shown in fig. 3, the step of determining the matching image specifically includes:

step 310, determining a matching score of the circuit layout pattern in the first image and the circuit layout pattern in the image to be tested.

In step 320, if the matching score is greater than the first threshold, the first image is determined to be a matching image.

Specifically, by acquiring the characteristics of the circuit layout pattern in the first image and the characteristics of the circuit layout pattern in the image to be detected, and then calculating the matching score of the two characteristics, the larger the matching score is, the more consistent the characteristics of the circuit layout pattern in the first image and the characteristics of the circuit layout pattern in the image to be detected are, the higher the similarity between the circuit layout pattern in the first image and the circuit layout pattern in the image to be detected is.

Based on this, when the matching score of a certain first image in the image library and the image to be measured is greater than the first threshold value, the first image can be determined as a matching image. The first threshold can be considered to be set according to actual requirements, and the empirical value of the first threshold is generally 0.96-1, and the specific value is preferably 0.98.

Since when more first images are stored in the image library, for example, 50 first images are stored in the image library, and the hot spot types and positions in part of the first images are the same, the process of sequentially calculating the matching score of each image and the image to be tested is cumbersome and time-consuming, in some possible implementations, in order to quickly complete the process of determining the matching images, the first images may be classified, and then step 220 is performed.

Specifically, after storing a plurality of first images including the circuit layout in the image library, the first images may be further divided into at least one image set, i.e., the plurality of first images may be divided into several types according to a specific classification basis. When the matching image is subsequently determined, the matching image can be quickly determined among the determined types.

The first image may be divided according to a type of the hotspot and/or based on a region around the hotspot in the circuit layout.

When dividing the first image according to the type of the hot spot, the following steps are required to be performed before dividing the first image:

the type of the hot spot in the first image is determined based on a difference between the area of the circuit layout and the actual area of the circuit layout in the first image.

If the difference between the area of the circuit layout and the actual area is the second preset value, and the area of the circuit layout is larger than the actual area, the hot spot type is a line bridging type, and the hot spot of the line bridging type is shown as a and b in fig. 4, where the hot spot of the line bridging type refers to that the two photoresist lines should be separated, the photoresist lines are not completely separated, and may also be colloquially described as "lotus root broken wire connection" between the photoresists. The impact of such hot spots is extremely large, the electrical properties are completely altered, and most often short circuits are the case.

If the difference between the area of the circuit layout and the actual area is the first preset value and the area of the circuit layout is smaller than the actual area, the hot spot type is a wire break type, the hot spot of the wire break type is shown as c and d in fig. 5, the hot spot of the wire break type refers to the situation that the photoresist wire breaks, and the influence of the hot spot is mostly circuit break.

If the difference between the area of the circuit layout and the actual area is a third preset value, the hot spot type is a contact hole type, wherein the third preset value is smaller than the second preset value and the third preset value is smaller than the first preset value, the hot spot of the contact hole type is shown as e in fig. 6, and the hot spot of the contact hole type is caused by incomplete contact hole and is very easy to cause short circuit of the circuit.

It should be noted that, the values of the first preset value, the second preset value and the third preset value are generally selected according to the actual scene. As an example, the first preset value has a value of 0.05, the second preset value has a value of 0.06, and the third preset value has a value of 0.03.

In some possible implementations, when the first image is divided based on the area around the hot spot in the circuit layout, as shown in fig. 7, the following steps are performed:

in step 710, a first region is determined by capturing a region around a location of a hot spot in a first image.

In step 720, it is determined that the first images with the same first region are the same type of image set.

It should be noted that, although the pattern structures of the hot spots are different, the first image may be divided by classifying the topological structure features of the circuit layout pattern. In the circuit layout, the topology structure describes the association between the hot spot and the surrounding area, and does not concern the details in the whole circuit layout graph or the proportional relationship among the areas in the circuit layout graph.

As an example, as shown in fig. 8A-8D, fig. 8A-8D are each a first image, the area indicated by M in fig. 8A-8D is a topology including a hot spot and an area around the hot spot, that is, M is a first area, the first images may be classified according to the shape of the first area, e.g., the fig. 8A-8D are divided into two different groups according to the first area M, where the first area M in fig. 8A and 8B are the same, and thus 8A and 8B are determined as the same type of image set; the first region M in fig. 8C and 8D is the same, and thus 8C and 8D are determined as the same type of image set.

It can be appreciated that after the first image is divided sequentially according to the type of the hot spot and based on the region around the hot spot in the circuit layout, the first image can be more accurately divided into different image sets, so that convenience in the process of acquiring the matching image of the image to be detected in the later stage is further improved.

In step 230, feature vectors of the matching image are obtained.

Each matching image has its own attribute, different attributes are represented by different attribute values, and a plurality of attribute values can be combined together to represent a vector, which is a feature vector (feature vector). That is, the feature vector refers to a set of attributes, and the feature vector of each matching image is attached to one matching image and is used to characterize the matching image.

In order to accurately characterize the matching image, the feature vector is a multi-dimensional vector, such as a two-dimensional vector, a four-dimensional vector, or a five-dimensional vector, that is, as shown in fig. 9, the position of the hot spot in the matching image B is indicated at a reference symbol a, and the feature vector may at least include any two of the following parameters: the shortest distance of the center of the hotspot to the edge of the matching image (e.g., d1 in fig. 9), the straight line shortest distance of the center of the hotspot to the edge of the circuit layout (e.g., d2 in fig. 9), the longest distance of the center of the hotspot to the edge of the circuit layout (e.g., d3 in fig. 9), the shortest distance of the edge of the matching image to the edge of the circuit layout (e.g., d4 in fig. 9), the longest distance of the edge of the matching image to the edge of the circuit layout (e.g., d5 in fig. 9), and the shortest distance of the center of the hotspot to the 4 corners of the circuit layout (e.g., d6 in fig. 9).

As an example, when the parameters of the feature vector include all the above parameters, the feature vector is a six-dimensional parameter, which may be expressed as p= (d 1, d2, d3, d4, d5, d 6), and at this time, the feature vector P can characterize the entire matching image.

After determining the feature vector, a pre-trained detection model may be adopted to predict the image to be detected corresponding to the feature vector, so as to obtain the type and the position of the hot spot in the image to be detected, namely, the following step 240 is performed:

In step 240, the feature vector is input into a predetermined inspection model, and the type and position of the hot spot in the wafer are determined based on the output of the inspection model.

The preset detection model is obtained through training a sample image of a known hot spot, specifically, as shown in fig. 10, the training process of the preset detection model includes:

step 1010, sequentially inputting the feature vectors of each sample image into a detection model to be trained, and predicting the sample hot spot positions and the sample hot spot types in each sample image through the detection model to be trained.

The size information of the sample image may be the same as or different from the size information of the first image, which is not limited in the embodiment of the present application.

And step 1020, performing iterative training on the detection model to be trained according to the real hot spot position, the sample hot spot position, the real hot spot type and the sample hot spot type, and obtaining the detection model after training is finished.

The detection model to be trained can adopt A Neural Network (ANN) detection model, the ANN refers to a complex network structure formed by interconnecting a large number of processing units (neurons), the ANN is a certain abstraction, simplification and simulation of human brain tissue structure and operation mechanism, the neural activity is simulated by a mathematical model, and the ANN detection model is an information processing system established based on the imitation of the brain neural network structure and function. The artificial neural network achieves the aim of processing the relation between information and analog input and output through repeated learning training of known information and a method of gradually adjusting and changing the connection weight of neurons.

In addition, the iteration number can be set according to actual requirements, for example, 10 iterations, 13 iterations, 15 iterations, etc., and the embodiment of the present application does not limit the type and the iteration number of the specific detection model to be trained.

In the above process, as shown in fig. 11, the initial detection model to be trained is trained by taking the sample images of the abundant known hot spots as the training data set, in the training process, the detection model to be trained extracts the rules in the training data set to obtain the target result (i.e., the sample hot spot position and the sample hot spot type), then the parameters of the detection model to be trained are adjusted according to the difference value between the real hot spot position and the sample hot spot position and the difference value between the real hot spot type and the sample hot spot type, and the parameters of the detection model to be trained are repeatedly trained, so that the errors between the real hot spot position and the sample hot spot position and the errors between the real hot spot type and the sample hot spot type output by the detection model to be trained are smaller than the preset deviation threshold value, at this time, training can be finished to obtain the trained detection model, i.e., the preset detection model, and the preset detection model can accurately predict the hot spot type and the position in the image to be tested based on the input feature vector.

It should be noted that, when the first image is divided into at least one image set, the preset detection model includes at least one classifier, and one image set corresponds to one classifier. And when the detection model to be trained is trained, the feature vector of the sample image is input into the detection model to be trained containing the corresponding classifier. When the feature vector of the matched image is input into a preset detection model, the feature vector is input into a classifier corresponding to the image set where the matched image is located. At this time, the different types of sample images are regarded as a single classification problem by the corresponding classifier, and a plurality of classifiers in the preset detection model are combined into a combined classifier, and the result of the combined classifier represents the result of the whole preset detection model.

It can be appreciated that when the preset detection model needs to identify a new hot spot, the classifier can be added or updated only in the preset detection model to adjust, and the whole process is simple and convenient.

In some possible implementation manners, because the circuit layout patterns in the first image are mostly patterns with different shapes, and the hot spot only occupies a small part of the area in the first image pattern, in order to further improve the accuracy of the preset detection model, when the detection model to be trained is trained, a certain amount of non-hot spot data can be input into the detection model to be trained. Specifically, second images with the same number as the first images are input into a detection model to be trained, wherein the second images are images which do not contain hot spots, the images contain circuit layouts, and the circuit layout figures in the first images can be the same as or different from the circuit layout figures in the second images. At this time, the hot spot and non-hot spot counting ratio in the training data set is relatively balanced, the to-be-trained detection model can fully learn and identify hot spots and non-hot spots (namely, background), accuracy of the detection model in the training stage is improved, and further the preset detection model can accurately detect the hot spot type and position in the image to be detected.

As shown in fig. 11, according to the above technical solution, a plurality of first images are stored in an image library after being divided into different types of image sets, so that a matching image matched with an image to be detected can be found out in the image library quickly, feature vectors of the matching image are input into a preset detection model, the feature vectors are detected through the preset detection model, the hot spot position in the detection image is determined, in the whole process, the feature vectors of the matching image can be directly input into the preset detection model for detection, the preset detection model does not need to extract features in each image to be detected, the data processing amount is reduced, and the final detection result is accurate. Compared with a method for detecting hot spots based on lithography simulation, the hot spot detection scheme provided by the application has the advantages of low calculation complexity and less time consumption on the basis of ensuring high accuracy.

Because the number of the first images in the image library is limited, and the types of the first images are limited, there are cases that no matching images exist in the image library, so in some possible implementation manners, as shown in fig. 11, if no matching images matched with the images to be detected exist in the image library, the feature vectors of the images to be detected are input into a preset detection model, and the detection images output by the detection model are obtained; and storing the detected image into an image library. In the process, the detection model can identify the type and the position of the hot spot in the image to be detected based on the feature vector, and the image to be detected is stored into the image library as the first image, so that the comprehensiveness of the image library is increased, and the subsequent hot spot detection process is facilitated.

In this embodiment, the process of determining the matching image is combined with the preset detection model, and for the case that the matching image exists, the feature vector of the matching image is input into the preset detection model, and at this time, the preset detection model does not need to perform complex feature extraction on the image to be detected, so that the calculation processing amount of the matching image can be greatly simplified. For the situation that no matching image exists, the preset detection model can effectively judge the position and the type of the hot spot in the image to be detected based on the feature vector of the image to be detected, namely, the scheme in the application has good prediction capability under the situation that no matching image exists, and has a wider application scene.

Based on the above hot spot detection method, the same technical concept is adopted, and the embodiment of the application also provides a hot spot detection device for implementing the hot spot detection method. The implementation of the solution provided by the device is similar to the implementation described in the above method embodiments.

In an exemplary embodiment, as shown in fig. 12, a hot spot detection apparatus in the present application includes:

the image storage module is used for storing a plurality of first images comprising circuit layouts into an image library, and the circuit layouts comprise hot spots;

The image matching module is used for determining a matching image matched with an image to be detected in the image library, wherein the first image comprises the matching image, and the image to be detected comprises a wafer needing hot spot detection;

the feature vector acquisition module is used for acquiring the feature vector of the matched image;

and a detection module: the method is used for inputting the feature vector into a preset detection model, and determining the type and the position of the hot spot in the wafer based on the output of the detection model.

It will be appreciated that the feature vector includes at least any two of the following parameters: the shortest distance from the center of the hot spot to the edge of the matching image, the straight line shortest distance from the center of the hot spot to the edge of the circuit layout, the longest distance from the center of the hot spot to the edge of the circuit layout, the shortest distance from the edge of the matching image to the edge of the circuit layout, the longest distance from the edge of the matching image to the edge of the circuit layout, and the shortest distance from the center of the hot spot to the 4 corners of the circuit layout.

It should be noted that, the training process of the preset detection model includes:

acquiring a plurality of sample images and feature vectors corresponding to the sample images, wherein the sample images comprise real hot spot positions and real hot spot types;

sequentially inputting the feature vectors of each sample image into a detection model to be trained, and predicting the sample hot spot positions and the sample hot spot types in each sample image through the detection model to be trained;

And carrying out iterative training on the detection model to be trained according to the real hot spot position, the sample hot spot position, the real hot spot type and the sample hot spot type, and obtaining a preset detection model after training is finished.

In some possible implementations, the preset detection model includes at least one classifier, and after storing a plurality of first images including the circuit layout in the image library, the apparatus further includes an image dividing module, where the dividing module is configured to:

dividing the first image into at least one image set, one image set corresponding to each classifier; at this time, the feature vector is input to a preset detection model, and the detection module includes a detection unit, where the detection unit is configured to input the feature vector to a classifier corresponding to an image set where the matching image is located.

In some possible implementations, when dividing the first image into at least one image set, the dividing module includes a dividing unit configured to:

and dividing the first image according to the type of the hot spot and/or based on the area around the hot spot in the circuit layout.

Specifically, before dividing the first image according to the type of the hot spot, the dividing unit is used for determining the type of the hot spot in the first image based on the difference between the area of the circuit layout and the actual area of the circuit layout in the first image;

If the difference value between the area of the circuit layout and the actual area is a first preset value, and the area of the circuit layout is smaller than the actual area, the hot spot type is a wire breakage type;

if the difference value between the area of the circuit layout and the actual area is a second preset value, and the area of the circuit layout is larger than the actual area, the hot spot type is a line bridging type;

if the difference between the area of the circuit layout and the actual area is a third preset value, the hot spot type is a contact hole type, wherein the third preset value is smaller than the second preset value, and the third preset value is smaller than the first preset value.

Dividing a first image based on the region around the hot spot in the circuit layout, wherein the dividing unit is used for determining the first region by intercepting the region around the hot spot position in the first image; the first images with the same first area are determined to be the same type of image set.

Obtaining a matching image matched with an image to be detected in an image library, wherein the image matching module comprises an image matching unit, and the image matching unit is used for:

determining the matching score of the circuit layout graph in the first image and the circuit layout graph in the image to be detected; and if the matching score is greater than a first threshold value, determining the first image as a matching image.

In some possible implementation manners, if no matching image matching the image to be detected exists in the image library, the detection module includes a direct detection unit, and the direct detection unit is used for inputting the feature vector of the image to be detected into a preset detection model to obtain a detection image output by the detection model; and storing the detected image into an image library.

It should be noted that, for specific limitation of the hot spot detection device, reference may be made to the limitation of the hot spot detection method hereinabove, and the description thereof will not be repeated here.

It should be noted that, each module in the above-mentioned hot spot detection device may be implemented in whole or in part by software, hardware, and a combination thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.

Furthermore, it will be appreciated by those skilled in the art that implementing all or part of the above-described embodiment method may be implemented by a computer program to instruct related hardware, where the computer program may be stored in a non-volatile computer readable storage medium, and the computer program may include the flow of the embodiment of the above-described hot spot detection method when executed.

Any reference to memory, storage, database, or other medium used in embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, or the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory. By way of illustration, and not limitation, RAM can be in the form of a variety of forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), and the like.

The foregoing examples represent only a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the invention. It should be noted that any modifications, equivalent substitutions, improvements, etc. made by those skilled in the art without departing from the spirit of the present application are intended to be included in the protection scope of the embodiments of the present application.

Claims (10)

1. A method of hotspot detection, comprising:

Storing a plurality of first images comprising circuit layouts into an image library, wherein the circuit layouts comprise hot spots, and the hot spots in at least two first images are different;

determining a matching image matched with an image to be detected in the image library, wherein the first image comprises the matching image, and the image to be detected comprises a wafer needing hot spot detection;

acquiring a feature vector of the matched image;

and inputting the feature vector into a preset detection model, and determining the type and the position of the hot spot in the wafer based on the output of the detection model.

2. The hot spot detection method according to claim 1, wherein the detection model includes at least one classifier, and wherein after storing the plurality of first images including the circuit layout in the image library, the method further comprises:

dividing the first image into at least one image set, one of the image sets corresponding to one of the classifiers;

the inputting the feature vector into a preset detection model comprises the following steps:

and inputting the feature vector into the classifier corresponding to the image set where the matched image is located.

3. The hot spot detection method according to claim 2, wherein the dividing the first image into at least one image set comprises:

And dividing the first image according to the type of the hot spot and/or based on the area around the hot spot in the circuit layout.

4. A hot spot detection method according to claim 3, wherein before dividing the first image according to the type of the hot spot, the method further comprises:

determining a hotspot type in the first image based on a difference between the area of the circuit layout and the actual area of the circuit layout in the first image;

if the difference value between the area of the circuit layout and the actual area is a first preset value, and the area of the circuit layout is smaller than the actual area, the hot spot type is a wire breakage type;

if the difference value between the area of the circuit layout and the actual area is a second preset value, and the area of the circuit layout is larger than the actual area, the hot spot type is a line bridging type;

and if the difference value between the area of the circuit layout and the actual area is a third preset value, the hot spot type is a contact hole type, wherein the third preset value is smaller than the second preset value, and the third preset value is smaller than the first preset value.

5. The method of detecting a hot spot according to claim 3, wherein the dividing the first image based on a region around the hot spot in the circuit layout includes:

determining a first area by intercepting areas around a hot spot position in the first image;

and determining that the first images with the same first area are the same type of image set.

6. The hot spot detection method according to claim 1, wherein the feature vector includes at least any two of the following parameters:

the shortest distance from the center of the hot spot to the edge of the matching image, the straight line shortest distance from the center of the hot spot to the edge of the circuit layout, the longest distance from the center of the hot spot to the edge of the circuit layout, the shortest distance from the edge of the matching image to the edge of the circuit layout, the longest distance from the edge of the matching image to the edge of the circuit layout and the shortest distance from the center of the hot spot to the 4 corners of the circuit layout.

7. The method for detecting a hot spot according to claim 1, wherein the obtaining the matching image in the image library, which matches the image to be detected, includes:

Determining the matching score of the circuit layout graph in the first image and the circuit layout graph in the image to be detected;

and if the matching score is greater than a first threshold value, determining the first image as the matching image.

8. The hotspot detection method of claim 1, further comprising:

if no matching image matched with the image to be detected exists in the image library, inputting the feature vector of the image to be detected into a preset detection model, and obtaining a detection image output by the detection model;

and storing the detection image into the image library.

9. The hot spot detection method according to any one of claims 1 to 8, wherein the training process of the detection model includes:

acquiring a plurality of sample images and feature vectors corresponding to the sample images, wherein the sample images comprise real hot spot positions and real hot spot types;

sequentially inputting the feature vectors of the sample images into a detection model to be trained, and predicting the sample hot spot positions and the sample hot spot types in the sample images through the detection model to be trained;

and carrying out iterative training on the detection model to be trained according to the real hot spot position, the sample hot spot position, the real hot spot type and the sample hot spot type, and obtaining the detection model after training is finished.

10. A hotspot detection apparatus, comprising:

the image storage module is used for storing a plurality of first images comprising circuit layouts into an image library, wherein the circuit layouts comprise hot spots;

the image matching module is used for determining a matching image matched with an image to be detected in the image library, wherein the first image comprises the matching image, and the image to be detected comprises a wafer needing hot spot detection;

the feature vector acquisition module is used for acquiring the feature vector of the matched image;

and a detection module: and the feature vector is used for inputting the feature vector into a preset detection model, and the type and the position of the hot spot in the wafer are determined based on the output of the detection model.

Images