A method and system for rapid and accurate image recognition of the bursting damage of active damage elements

Abstract

The present application relates to a method and system for rapid and accurate image recognition of the bursting damage of active damage elements. The method includes: in response to the obtained captured image, performing grayscale processing on the captured image to obtain a reference damage image; dividing the regions on the reference damage image to obtain damage regions and non-damage regions; dividing the damage regions according to regional features to obtain main regions and sub-regions, where the regional features include shape and color; determining the attribution relationship between the main regions and the sub-regions; classifying the main regions according to the attribution relationship of the sub-regions to obtain main bursting damage regions and main non-bursting damage regions. The method and system for rapid and accurate image recognition of the bursting damage of active damage elements disclosed in the present application can quickly divide and classify the regions on the captured image, and then determine the bursting damage region, providing data reference for subsequent adjustments.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and particularly relates to a method and system for rapid and accurate image recognition of the bursting damage of active damage elements. Background Art

[0002] The development stage of damage elements is mainly divided into two stages: using inert materials and using active materials. Inert material damage elements mainly cause damage to the target by mechanical damage, while active material damage elements use both mechanical damage and bursting damage to cause damage to the target.

[0003] The core difference between the two lies in the different materials used. Active materials have both mechanical strength similar to metals and chemical energy equivalent to high-energy explosives. After hitting the target, they can cause primary damage to the weak parts on the surface of the target and secondary damage after entering the target interior.

[0004] During test evaluation and actual use, it is necessary to evaluate the use effect of active damage elements. The purpose of the evaluation is to determine the ratio of mechanical damage and bursting damage, and then make targeted adjustments. The current method used is image processing, but due to the small spacing of the damage areas, there is a certain difficulty in recognition. Summary of the Invention

[0005] This application provides a method and system for rapid and accurate image recognition of the bursting damage of active damage elements, which can quickly divide and classify the areas on the captured image, and then determine the bursting damage area to provide data reference for subsequent adjustments.

[0006] The above object of this application is achieved through the following technical solutions:

[0007] In a first aspect, this application provides a method for rapid and accurate image recognition of the bursting damage of active damage elements, including:

[0008] In response to the obtained captured image, perform grayscale processing on the captured image to obtain a reference damage image;

[0009] Divide the areas on the reference damage image to obtain damage areas and non-damage areas;

[0010] Divide the damage areas according to regional characteristics to obtain main areas and sub-areas, and the regional characteristics include shape and color;

[0011] Determine the attribution relationship between the main areas and the sub-areas;

[0012] Classify the main areas according to the attribution relationship of the sub-areas to obtain main bursting damage areas and main non-bursting damage areas;

[0013] Among them, when a sub-region has an associated relationship with at least two main regions, the classification of the main region is determined according to the edge characteristics of the main region.

[0014] In a possible implementation manner of the first aspect, dividing the regions on the reference damage image and obtaining the damaged regions and non-damaged regions includes:

[0015] Using an edge extraction method to divide the reference damage image to obtain multiple feature regions;

[0016] Classifying the feature regions according to the shape characteristics, and the feature regions of the same type have similar shape characteristics;

[0017] Dividing the classified feature regions into damaged regions and non-damaged regions;

[0018] Among them, the feature regions belonging to the damaged regions have similar shapes and / or colors; the non-damaged regions are the parts of the reference damage image that do not belong to the damaged regions.

[0019] In a possible implementation manner of the first aspect, classifying the feature regions according to the shape characteristics includes:

[0020] Determining the edge of the feature region and analyzing the edge characteristics of the feature region, and the edge characteristics include shape change characteristics and color change characteristics;

[0021] Classifying the feature regions according to the edge characteristics of the feature region, and the classification includes mechanical damage type and explosion damage type.

[0022] In a possible implementation manner of the first aspect, analyzing the shape change characteristics and color change characteristics in the edge characteristics of the feature region includes:

[0023] Randomly establish an analysis reference line on the feature region, and both ends of the analysis reference line are located outside the feature region;

[0024] Calculate the color change trend on the analysis reference line and determine the rotation direction according to the color change trend on the analysis reference line, or randomly rotate the analysis reference line and then calculate the color change trend on the analysis reference line;

[0025] Determine the change points on the analysis reference line;

[0026] Assign types to the change points, and the change point types include shape change and color change;

[0027] Take the change points of the shape change type as the shape change characteristics of the edge characteristics, and take the change points of the color change type as the color change characteristics of the edge characteristics.

[0028] In a possible implementation of the first aspect, determining the change points on the analysis reference line includes:

[0029] Using the pixel points on the analysis reference line to establish an analysis feature curve;

[0030] Transferring the analysis feature curve into the time domain and frequency domain for decomposition to obtain a set of analysis feature curves;

[0031] Determining the change points according to the starting position points and ending position points of the analysis feature curves in the set of analysis feature curves.

[0032] In a possible implementation of the first aspect, assigning change point types includes:

[0033] Determining the color difference values on both sides of the change point;

[0034] When the color difference values on both sides of the same change point are within the allowable range, the change point type is shape change; when the color difference values on both sides of the same change point exceed the allowable range, the change point type is color change.

[0035] In a possible implementation of the first aspect, when determining the attribution relationship between the main region and the sub-region, the used attribution relationship is inclusion.

[0036] In a second aspect, the present application provides a fast and accurate image recognition device for the bursting damage of active damage elements, including:

[0037] An image processing unit, configured to perform gray processing on the captured image in response to the obtained captured image to obtain a reference damage image;

[0038] A first region division unit, configured to divide the region on the reference damage image to obtain a damaged region and a non-damaged region;

[0039] A second region division unit, configured to divide the damaged region according to the region features to obtain a main region and a sub-region, where the region features include shape and color;

[0040] An attribution relationship determination unit, configured to determine the attribution relationship between the main region and the sub-region;

[0041] A region classification unit, configured to classify the main region according to the attribution relationship of the sub-region to obtain a main region of bursting damage and a main region of non-bursting damage;

[0042] Wherein, when a sub-region has an association relationship with at least two main regions, the classification of the main region is determined according to the edge features of the main region.

[0043] In a third aspect, the present application provides a fast and accurate image recognition system for the bursting damage of active damage elements, and the system includes:

[0044] One or more memories for storing instructions; and

[0045] One or more processors for invoking and running the instructions from the memory and performing the method as described in the first aspect and any possible implementation manners of the first aspect.

[0046] In a fourth aspect, the present application provides a computer-readable storage medium, which includes:

[0047] A program, when the program is run by a processor, the method as described in the first aspect and any possible implementation manners of the first aspect is executed.

[0048] In a fifth aspect, the present application provides a computer program product, including program instructions, when the program instructions are run by a computing device, the method as described in the first aspect and any possible implementation manners of the first aspect is executed.

[0049] In a sixth aspect, the present application provides a chip system, which includes a processor for implementing the functions involved in the above aspects, for example, generating, receiving, sending, or processing the data and / or information involved in the above method.

[0050] The chip system may be composed of chips or may include chips and other discrete devices.

[0051] In a possible design, the chip system further includes a memory for storing necessary program instructions and data. The processor and the memory may be decoupled and disposed on different devices and connected by a wired or wireless manner, or the processor and the memory may also be coupled on the same device. Description of the Drawings

[0052] Figure 1 is a schematic block diagram of the step flow of a method for fast and accurate image recognition of the burst damage of an active damage element provided by the present application.

[0053] Figure 2 is a schematic diagram of a damaged area and a non-damaged area provided by the present application.

[0054] Figure 3 is a schematic diagram of a main area and a sub-area provided by the present application.

[0055] Figure 4 is a schematic diagram of establishing an analysis reference line provided by the present application.

[0056] Figure 5 is a schematic diagram of rotating an analysis reference line provided by the present application.

[0057] Figure 6 This is a schematic diagram of the principle for obtaining change points provided by this application. Specific implementation manners

[0058] The following further elaborates on the technical solutions in this application with reference to the accompanying drawings.

[0059] This application discloses a fast and accurate image recognition method for the burst damage of active damage elements. Please refer to Figure 1 , in some examples, the fast and accurate image recognition method for the burst damage of active damage elements disclosed in this application includes the following steps:

[0060] S101. In response to the obtained captured image, perform grayscale processing on the captured image to obtain a reference damage image.

[0061] S102. Divide the regions on the reference damage image to obtain damage regions and non-damage regions.

[0062] S103. Divide the damage regions according to region features to obtain main regions and sub-regions, where the region features include shape and color.

[0063] S104. Determine the attribution relationship between the main regions and the sub-regions.

[0064] S105. Classify the main regions according to the attribution relationship of the sub-regions to obtain main burst damage regions and main non-burst damage regions.

[0065] Among them, when a sub-region has an association relationship with at least two main regions, determine the classification of the main regions according to the edge features of the main regions.

[0066] First, introduce the relevant content of the active damage elements. The active damage elements involved in this application are deployed at the warhead of the weapon. After the warhead explodes, kinetic energy is imparted to the active damage elements. At this time, the active damage elements approach the target. After coming into contact with the target, the active damage elements damage the target surface. The damage methods here are mechanical damage and explosion damage. Some active damage elements will enter the target surface after passing through the target surface to cause secondary damage, and at this time, it is generally explosion damage.

[0067] The prerequisite for generating mechanical damage is that the skin thickness of the target surface is relatively thin. Therefore, active damage elements are mostly used to strike aircraft, and in the test process, the test objects are also mostly aircraft.

[0068] Returning to the technical solution provided by this application, in step S101, in response to the obtained captured image, perform grayscale processing on the captured image to obtain a reference damage image. Here, the captured image is obtained by an image sensor at the test site, or an image sensor carried on the weapon, or by a combat unit that launches the weapon.

[0069] After obtaining the reference damage image, perform step S102. Please refer to Figure 2 , in this step, the area on the reference damage image is divided to obtain the damaged area and the undamaged area. The damaged area refers to the area where the active damage element damages the target surface, and the undamaged area refers to the area on the target surface that is not damaged by the active damage element. One purpose of dividing the damaged area and the undamaged area is to determine the area of the damaged area, and the other purpose is to facilitate the subsequent division of the main bursting damage area and the main non-bursting damage area.

[0070] In step S103, the damaged area will be divided according to the regional characteristics. Please refer to Figure 3 , to obtain the main area and the sub-area. The main area is directly caused by the active damage element, and the sub-area is indirectly caused by the active damage element. In this application, the sub-area is generated by the active damage element during the explosion.

[0071] The main area here refers to the mechanical damage type and the explosion damage type in the damaged area, and the sub-area refers to the explosion damage derivative type in the damaged area.

[0072] In step S104, it is necessary to determine the attribution relationship between the main area and the sub-area. Combining the content in the previous text, it can be seen that the attribution relationship here refers to that the reason for the appearance of the sub-area is because of the main area, that is, the two belong to the same active damage element.

[0073] In step S105, the main area will be classified according to the attribution relationship of the sub-area to obtain the main bursting damage area and the main non-bursting damage area. Combining the content mentioned in the previous text, it can be seen that for the main area with a sub-area, it is designated as the main bursting damage area, and vice versa as the main non-bursting damage area.

[0074] Of course, at this time, there may be a situation where a sub-area is associated with at least two main areas. At this time, it is necessary to determine the classification of the main area according to the edge characteristics of the main area. This part of the content will be further described in the subsequent content.

[0075] In some possible implementation manners, when determining the attribution relationship between the main area and the sub-area, the used attribution relationship is inclusion. Here, the definition of inclusion is that the main area is located inside the sub-area, or a part of the main area is located inside the sub-area. Generally, the requirement here is that at least 70% of the edge of the main area is located inside the sub-area.

[0076] Generally speaking, through the above methods, the main areas of burst damage and non-burst damage on the target surface can be quickly determined. After determination, subsequent analysis work can be carried out. For example, at this time, quantitative analysis work can be carried out. For example, by calculating the ratio of the main areas of burst damage and non-burst damage, parameters such as the impact speed and flight distance of the active damage elements can be judged whether they are appropriate.

[0077] In some examples, the specific method of dividing the area on the reference damage image to obtain the damaged area and the non-damaged area is as follows:

[0078] S201, use the edge extraction method to divide the reference damage image to obtain multiple feature regions;

[0079] S202, classify the feature regions according to the shape features. The feature regions of the same type have similar shape features;

[0080] S203, divide the classified feature regions into damaged areas and non-damaged areas;

[0081] Among them, the feature regions belonging to the damaged area have similar shapes and / or colors; the non-damaged area is the part of the reference damage image that does not belong to the damaged area.

[0082] In step S201, first use the edge extraction method to divide the reference damage image. The edge extraction method is implemented using edge detection algorithms, including Sobel operator, Prewitt operator, Roberts operator, and Canny operator, etc.

[0083] Then in step S202, classify the feature regions according to the shape features. The feature regions of the same type have similar shape features. Specifically, the shape of the impact-type damage is similar to the shape of the active damage element. Here, it is necessary to collect the impact-type damage generated by different angles of the active damage element and establish a database, that is, the impact-type damage can be determined by comparison processing. The characteristic of the explosion-type damage is that it has irregular edges. After using the edge detection algorithm to extract the edges, the enclosed area of the irregular edges is used as the damaged area.

[0084] From Figure 2 it can be seen that the damaged area has obvious damage characteristics. The recognition method of this damage characteristic is as follows:

[0085] The characteristic regions of the mechanical damage type have the characteristic that their shapes are similar to the shape of the active damage element, because the characteristic regions of the mechanical damage type are generated by impact and the edges are relatively regular. The explosion damage type has the characteristic that its shape edges are significantly irregular.

[0086] The core of these two characteristic regions lies in whether the edge is regular. One way to determine whether the edge is regular is to first segment the edge of a characteristic region, because the edge of a characteristic region generally consists of multiple segments.

[0087] Then calculate the regularity and irregularity of the edge. The specific method is to establish a straight line on the segmented edge, and the straight line needs to be as parallel as possible to the corresponding edge. For example, a rectangular region can be obtained based on the edge, and then the midline of the rectangular region is used as a reference for establishing the straight line, and a straight line established is parallel to the midline of the rectangular region.

[0088] Take multiple points at intervals on the segmented edge, and then calculate the minimum straight-line distance between these points and the straight line.

[0089] Statistical fluctuation range of these minimum straight-line distances. The fluctuation range is the difference between the maximum value and the minimum value of the minimum straight-line distance. When determining the maximum value and the minimum value, one or two maximum values and one or two minimum values need to be removed first.

[0090] After obtaining the fluctuation range, classify the edge of the characteristic region according to the fluctuation range. The classification method is to compare the fluctuation range with a set value. When the fluctuation range is on the left side (less than) of the set value, the region corresponding to the edge of the characteristic region is of the mechanical damage type, and vice versa is of the explosion damage type.

[0091] The explosion damage derivative type region is located around the explosion damage type region and also has an irregular edge.

[0092] The non-damage region is the part of the reference damage image that does not belong to the damage region. There is a color difference between the explosion damage derivative type region and the non-damage region. Here, the explosion damage derivative type region needs to be processed separately, that is, when obtaining the mechanical damage type region and the explosion damage type region, the surrounding regions of these two regions are included in the scope of investigation of the explosion damage derivative type region.

[0093] The edge of the explosion damage derivative type region is also irregular. Use the edge classification method described above for processing, and finally take the remaining region part as the non-damage region.

[0094] Finally, in step S203, classify the classified characteristic regions into damage regions and non-damage regions, as Figure 2 shown.

[0095] The function of determining the damage region area includes determining the damage range (damage degree) of the target surface, determining whether the deployment method and flight method of the active damage elements are appropriate, and determining whether the departure distance of the active damage elements is appropriate, etc. Of course, this is only an example here and does not limit this application.

[0096] The present application also provides another way to distinguish mechanical damage types and explosion damage types. This way is mainly used to distinguish mechanical damage types and explosion damage types that are both located in the explosion damage derivative area, where it is impossible to rely on the explosion damage derivative area for distinction.

[0097] The specific way to classify the feature area according to the shape feature is as follows:

[0098] S301, determine the edge of the feature area and analyze the edge features of the feature area. The edge features include shape change features and color change features;

[0099] S302, classify the feature area according to the edge features of the feature area. The classification includes mechanical damage types and explosion damage types.

[0100] In steps S301 and S302, first, it is necessary to determine the edge of the feature area and analyze the edge features of the feature area. There are two types of edge features: shape change features and color change features. Then, classify the feature area according to the edge features of the feature area. The classification includes mechanical damage types and explosion damage types.

[0101] Please refer to Figure 4 and Figure 5 , the specific ways to analyze the shape change features and color change features in the edge features of the feature area are as follows:

[0102] S401, randomly establish an analysis reference line on the feature area, and both ends of the analysis reference line are located outside the feature area;

[0103] S402, calculate the color change trend on the analysis reference line and determine the rotation direction according to the color change trend on the analysis reference line, or randomly rotate the analysis reference line and then calculate the color change trend on the analysis reference line;

[0104] S403, determine the change points on the analysis reference line;

[0105] S404, assign types to the change points. The change point types include shape change and color change;

[0106] S405, regard the change points of the shape change type as the shape change features of the edge features and regard the change points of the color change type as the color change features of the edge features.

[0107] In steps S401 to S405, first, a reference analysis line is randomly established on the feature region, and it is required that both ends of the reference analysis line are located outside the feature region. Then, the color change trend on the reference analysis line is calculated, and the rotation direction is determined according to the color change trend on the reference analysis line. The purpose of determining the rotation direction is to enable the reference analysis line to pass through the center position of the feature region. Figure 5 The dashed line in

[0108] When the color change trend cannot be determined, the reference analysis line is randomly rotated and then the color change trend on the reference analysis line is calculated.

[0109] Regarding the color change trend, it means that the colors on the reference analysis line are significantly different, rather than remaining unchanged.

[0110] Then, the change points on the reference analysis line are determined and the change point types are assigned. The change point types include shape change and color change. Finally, the change points of the shape change type are used as the shape change features of the edge feature, and the change points of the color change type are used as the color change features of the edge feature.

[0111] The specific method for determining the change points on the reference analysis line is as follows:

[0112] Use the pixel points on the reference analysis line to establish an analysis feature curve;

[0113] Transfer the analysis feature curve into the time domain and frequency domain for decomposition to obtain an analysis feature curve group;

[0114] Determine the change points according to the starting position points and ending position points of the analysis feature curves in the analysis feature curve group.

[0115] Please refer to Figure 6 , in the above method, the wavelet decomposition method is used to process the reference analysis line. At this time, an analysis feature curve group will be obtained. The curves in the analysis feature curve group have starting points and ending points. Statistics are performed on these starting points and ending points, and the convergence region of the starting points and ending points is the location where the change points are located.

[0116] When the obtained convergence region is a single point, directly use this point as the change point. When the obtained convergence region is multiple points, use a line segment to connect these points, and then use the midpoint of this line segment as the change point.

[0117] Then, it is necessary to determine the types of the change points. The specific method is as follows:

[0118] Determine the color difference between both sides of the change point;

[0119] When the color difference on both sides of the same change point is within the allowable range, the change point type is shape change. When the color difference on both sides of the same change point exceeds the allowable range, the change point type is color change.

[0120] For the areas of mechanical damage, the change points include a change point of color change type and a change point of shape change type. The corresponding situation is that the active damage element causes damage to the area, and there is bending deformation at the edge of the damage. Because of the bending deformation, there is a certain difference in brightness of the color in the same area, and the point at the junction of light and dark is the change point of color change type.

[0121] And the distance between these two change points is relatively short, which is a significant feature.

[0122] For the areas of explosion damage, the change points include two change points of color change type. The distance between these two change points of color change type is relatively large, which is also a significant feature. At the same time, there may also be change points of shape change type between these two change points of color change type, but this situation is not considered in this case.

[0123] By performing type determination and distance determination on the change points, the characteristic areas of mechanical damage and explosion damage can be effectively distinguished. For the distance value used in the distance determination, it is generally determined according to the data collected during the test. The distance in the first case is generally controlled within 2 - 5 mm, and the distance in the second case is generally controlled to be greater than 5 mm.

[0124] The present application also provides a fast and accurate image recognition device for the bursting damage of active damage elements, including:

[0125] An image processing unit, configured to perform gray processing on the captured image in response to the obtained captured image to obtain a reference damage image;

[0126] A first region division unit, configured to divide the regions on the reference damage image to obtain a damaged region and a non-damaged region;

[0127] A second region division unit, configured to divide the damaged region according to the region characteristics to obtain a main region and a sub-region, where the region characteristics include shape and color;

[0128] An attribution relationship determination unit, configured to determine the attribution relationship between the main region and the sub-region;

[0129] A region classification unit, configured to classify the main region according to the attribution relationship of the sub-region to obtain a main region of bursting damage and a main region of non-bursting damage;

[0130] Among them, when a sub-region has an associated relationship with at least two main regions, the classification of the main region is determined according to the edge characteristics of the main region.

[0131] Furthermore, dividing the regions on the reference damage image and obtaining damage regions and non-damage regions includes:

[0132] Using an edge extraction method to divide the reference damage image to obtain multiple feature regions;

[0133] Classifying the feature regions according to the shape characteristics, and the feature regions of the same type have similar shape characteristics;

[0134] Dividing the classified feature regions into damage regions and non-damage regions;

[0135] Among them, the feature regions belonging to the damage regions have similar shapes and / or colors; the non-damage regions are the parts of the reference damage image that do not belong to the damage regions.

[0136] Furthermore, classifying the feature regions according to the shape characteristics includes:

[0137] Determining the edges of the feature regions and analyzing the edge characteristics of the feature regions, and the edge characteristics include shape change characteristics and color change characteristics;

[0138] Classifying the feature regions according to the edge characteristics of the feature regions, and the classification includes mechanical damage type and explosion damage type.

[0139] Furthermore, analyzing the shape change characteristics and color change characteristics in the edge characteristics of the feature regions includes:

[0140] Randomly establishing an analysis reference line on the feature region, and both ends of the analysis reference line are located outside the feature region;

[0141] Calculating the color change trend on the analysis reference line and determining the rotation direction according to the color change trend on the analysis reference line or randomly rotating the analysis reference line and then calculating the color change trend on the analysis reference line;

[0142] Determining the change points on the analysis reference line;

[0143] Assigning types to the change points, and the change point types include shape change and color change;

[0144] Taking the change points of the shape change type as the shape change characteristics of the edge characteristics and taking the change points of the color change type as the color change characteristics of the edge characteristics.

[0145] Furthermore, determining the change points on the analysis reference line includes:

[0146] An analysis feature curve is established using the pixels on the analysis reference line.

[0147] The analysis feature curve is transferred into the time domain and the frequency domain for decomposition to obtain an analysis feature curve group.

[0148] Change points are determined based on the starting position points and ending position points of the analysis feature curves in the analysis feature curve group.

[0149] Further, the types assigned to the change points include:

[0150] Determine the color difference on both sides of the change point.

[0151] When the color difference on both sides of the same change point is within the allowable range, the change point type is shape change; when the color difference on both sides of the same change point exceeds the allowable range, the change point type is color change.

[0152] Further, when determining the attribution relationship between the main region and the sub-region, the attribution relationship used is inclusion.

[0153] In one example, the units in any of the above devices may be one or more integrated circuits configured to implement the above methods. For example: one or more application specific integrated circuits (ASICs), or, one or more digital signal processors (DSPs), or, one or more field programmable gate arrays (FPGAs), or a combination of at least two of these integrated circuit forms.

[0154] Again, when the units in the device can be implemented in the form of a processing element scheduler, the processing element may be a general-purpose processor, such as a central processing unit (CPU) or other processors that can call programs. Again, these units may be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0155] In this application, various objects such as various messages / information / devices / network elements / systems / devices / actions / operations / processes / concepts, etc. may be named. It can be understood that these specific names do not constitute limitations on the relevant objects, and the assigned names may change with factors such as the scenario, context, or usage habits. The understanding of the technical meaning of the technical terms in this application should be mainly determined from the functions and technical effects reflected / executed in the technical solution.

[0156] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0157] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0158] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0159] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0160] It should also be understood that in each embodiment of the present application, the first, second, etc. are only used to indicate that multiple objects are different. For example, the first time window and the second time window are only used to indicate different time windows. And it should not have any impact on the time window itself. The above first, second, etc. should not impose any restrictions on the embodiments of the present application.

[0161] It should also be understood that in each embodiment of the present application, if there is no special explanation and logical conflict, the terms and / or descriptions between different embodiments are consistent and can be mutually referred to, and the technical features in different embodiments can be combined to form new embodiments according to their internal logical relationships.

[0162] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a computer-readable storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned computer-readable storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0163] This application also provides a fast and accurate image recognition system for the blast damage of active damage elements. The system includes:

[0164] One or more memories for storing instructions; and

[0165] One or more processors for calling and running the instructions from the memory and executing the methods described in the above content.

[0166] This application also provides a computer program product. This computer program product includes instructions that, when executed, cause the terminal device and the network device to perform the operations of the terminal device and the network device corresponding to the above methods.

[0167] This application also provides a chip system. This chip system includes a processor for implementing the functions involved in the above content. For example, generating, receiving, sending, or processing the data and / or information involved in the above methods.

[0168] This chip system can be composed of chips or can also include chips and other discrete devices.

[0169] The processor mentioned anywhere above can be a CPU, a microprocessor, an ASIC, or an integrated circuit for controlling the execution of one or more programs of the above methods for transmitting feedback information.

[0170] In a possible design, this chip system further includes a memory for storing necessary program instructions and data. The processor and the memory can be decoupled and are respectively arranged on different devices and connected by wired or wireless means to support the chip system in implementing various functions in the above embodiments. Or, the processor and the memory can also be coupled on the same device.

[0171] Optionally, the computer instructions are stored in a memory.

[0172] Optionally, the memory is a storage unit within the chip, such as a register, cache, etc., and the memory can also be a storage unit outside the chip within the terminal, such as a ROM or other types of static storage devices that can store static information and instructions, a RAM, etc.

[0173] It can be understood that the memory in the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories.

[0174] The non-volatile memory can be a ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), or a flash memory.

[0175] The volatile memory can be a RAM, which is used as an external cache. There are various different types of RAM, such as a static random access memory (SRAM), a dynamic random access memory (DRAM), a synchronous DRAM (SDRAM), a double data rate SDRAM (DDR SDRAM), an enhanced SDRAM (ESDRAM), a synch link DRAM (SLDRAM), and a direct memory bus random access memory.

[0176] The embodiments of the specific implementation manners are all preferred embodiments of the present application, and do not limit the protection scope of the present application accordingly. Therefore, all equivalent changes made according to the structure, shape, and principle of the present application shall be covered within the protection scope of the present application.

Claims

1. A rapid and accurate image recognition method for the bursting damage of active damage elements, characterized in that, Including: In response to the obtained captured image, performing grayscale processing on the captured image to obtain a reference damage image; Dividing the regions on the reference damage image to obtain damage regions and non-damage regions; Dividing the damage regions according to regional features to obtain main regions and sub-regions, where the regional features include shape and color. The main regions are directly caused by active damage elements, including mechanical damage types and explosion damage types in the damage regions, and the sub-regions are indirectly caused by active damage elements and are explosion damage derivative types in the damage regions; Determining the attribution relationship between the main regions and the sub-regions; Classifying the main regions according to the attribution relationship of the sub-regions to obtain main burst damage regions and main non-burst damage regions. The main regions with sub-regions are main burst damage regions, and the main regions without sub-regions are main non-burst damage regions; Among them, when a sub-region has an associated relationship with at least two main regions, determining the classification of the main regions according to the edge features of the main regions.

2. The rapid and precise image recognition method for the blast damage of active damage elements according to claim 1, characterized in that, Dividing the regions on the reference damage image and obtaining damage regions and non-damage regions includes: Using an edge extraction method to divide on the reference damage image to obtain multiple feature regions; Classifying the feature regions according to shape features, and the feature regions of the same type have similar shape features; Dividing the classified feature regions into damage regions and non-damage regions; Among them, the feature regions belonging to the damage regions have similar shapes and / or colors; the non-damage regions are the parts of the reference damage image that do not belong to the damage regions.

3. The rapid and precise image recognition method for the bursting damage of active damage elements according to claim 2, characterized in that, Classifying the feature regions according to shape features includes: Determining the edges of the feature regions and analyzing the edge features of the feature regions, where the edge features include shape change features and color change features; Classifying the feature regions according to the edge features of the feature regions, and the classification includes mechanical damage types and explosion damage types.

4. The rapid and accurate image recognition method for the burst damage of the active damage element according to claim 3, characterized in that, Analyzing the shape change features and color change features in the edge features of the feature regions includes: Randomly establishing an analysis reference line on the feature region, and both ends of the analysis reference line are located outside the feature region; Calculating the color change trend on the analysis reference line and determining the rotation direction according to the color change trend on the analysis reference line, or randomly rotating the analysis reference line and then calculating the color change trend on the analysis reference line; Determining the change points on the analysis reference line; Assigning change point types, where the change point types include shape changes and color changes; Regarding the change points of the shape change type as the shape change features of the edge features and regarding the change points of the color change type as the color change features of the edge features.

5. The rapid and precise image recognition method for the bursting damage of the active damage element according to claim 4, characterized in that, Determining the change points on the analysis reference line includes: Using the pixel points on the analysis reference line to establish an analysis feature curve; Transferring the analysis feature curve into the time domain and frequency domain for decomposition to obtain an analysis feature curve group; Determining the change points according to the starting position points and ending position points of the analysis feature curves in the analysis feature curve group.

6. The rapid and precise image recognition method for the bursting damage of active damage elements according to claim 4, characterized in that, Assigning change point types includes: Determining the color difference between both sides of the change point; When the color difference between both sides of the same change point is within the allowable range, the change point type is a shape change. When the color difference between both sides of the same change point exceeds the allowable range, the change point type is a color change.

7. The rapid and precise image recognition method for the bursting damage of the active damage element according to claim 1, wherein When determining the attribution relationship between the main area and the sub - area, the attribution relationship used is inclusion.

8. A rapid and precise image recognition device for the bursting damage of active damage elements, characterized in that Including: An image processing unit, configured to perform grayscale processing on the captured image in response to the obtained captured image to obtain a reference damage image; A first area division unit, configured to divide the area on the reference damage image to obtain a damaged area and a non - damaged area; A second area division unit, configured to divide the damaged area according to area features to obtain a main area and a sub - area, where the area features include shape and color. The main area is directly caused by active damage elements and includes mechanical damage types and explosion damage types in the damaged area. The sub - area is indirectly caused by active damage elements and is the explosion - damage - derived type in the damaged area; An attribution relationship determination unit, configured to determine the attribution relationship between the main area and the sub - area; An area classification unit, configured to classify the main area according to the attribution relationship of the sub - area to obtain a main area of burst damage and a main area of non - burst damage. The main area with a sub - area is the main area of burst damage, and the main area without a sub - area is the main area of non - burst damage; Among them, when a sub - area has an association relationship with at least two main areas, the classification of the main area is determined according to the edge features of the main area.

9. A rapid and precise image recognition system for the bursting damage of active damage elements, characterized in that, The system includes: One or more memories, configured to store instructions; and One or more processors, configured to call and run the instructions from the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer - readable storage medium includes: A program, when the program is run by a processor, the method according to any one of claims 1 to 7 is executed.

Images