Electrical component wiring diagram recognition system using image segmentation techniques

Abstract

The present application belongs to the technical field of electrical component wiring identification, and particularly relates to an electrical component wiring diagram identification system using image segmentation technology. The system extracts original image quality indicators and image quality indicators after electrical component segmentation by means of historical identification records of electrical component wiring diagrams, thereby identifying sensitive image quality indicators of electrical components in view of fluctuation analysis of the image quality indicators, which can effectively improve the purpose of image processing and maximize the segmentation efficiency. At the same time, the system predicts the segmentation duration when the electrical components in the current electrical component wiring diagram are segmented, determines the initial segmentation order of the electrical components according to the prediction results, and then performs real-time system resource detection in the operation according to the initial segmentation order, and dynamically adjusts the initial segmentation order accordingly. This method can realize adaptive adjustment based on system resource constraints, so that the electrical component segmentation is more in line with the actual situation of system resources.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of electrical component wiring identification, and particularly relates to an electrical component wiring diagram identification system using image segmentation technology. BACKGROUND

[0002] With the increasing demand of consumers for electrical products, especially digital products, the production and assembly demand of electrical products has also increased significantly. In order to improve production efficiency and quality, modern manufacturing industry widely adopts automated production lines to use assembly robots for efficient assembly. In this background, it is particularly important to efficiently identify the electrical component wiring diagram. Through efficient identification of the electrical component wiring diagram, the automated production line can quickly and accurately complete the assembly task, reduce the production cycle, and improve the overall production efficiency.

[0003] In the electrical component wiring diagram, there are usually various electrical components. The identification of electrical components is actually to accurately segment each electrical component from the wiring diagram to determine its specific position in the circuit and assembly position. In the segmentation operation of electrical components, image quality has a key influence on the segmentation effect. Because high-quality images can provide more detailed information, reduce noise and distortion, and thus improve the execution efficiency of the segmentation algorithm. Therefore, image quality processing is often needed in the electrical component segmentation process. However, the existing image quality processing method is generally generalized and does not take into account the sensitivity of different electrical component segmentation to image quality indicators, resulting in a lack of pertinence in image quality processing. This generalized processing method is easy to waste a lot of time on image quality processing and may require secondary processing, thereby increasing the processing time and consumption of computing resources and reducing the overall segmentation efficiency.

[0004] In addition, in the existing electrical component segmentation process, a certain principle is usually fixed, that is, to preferentially segment complex electrical components to ensure the identification accuracy of these complex components and reduce false positives and false negatives in subsequent processing. However, the segmentation of complex electrical components often consumes a large amount of system resources. When system resources are abundant, this segmentation order will not cause problems. However, if the segmentation of complex electrical components is still blindly performed when system resources are limited, it is easy to cause segmentation interruption, which on the one hand causes the current task to be unable to be completed and needs to be restarted, increasing the processing time and resource consumption, and on the other hand, the segmentation interruption may cause partial data loss, which needs to be reacquired and processed, affecting the continuity and integrity of the overall work. SUMMARY

[0005] In view of this, the present application aims to provide an electrical component wiring diagram identification system using image segmentation technology, which can effectively solve the problems existing in the prior art.

[0006] The purpose of the present application can be realized by the following technical solutions: an electrical component wiring diagram recognition system using image segmentation technology, comprising the following modules: a historical recognition record calling module, used to call the historical recognition record of the electrical component wiring diagram, and extract the electrical components contained in the electrical component wiring diagram from the historical recognition record.

[0007] An image quality indicator extraction module is used to extract the image quality indicators of the original electrical component wiring diagram from the historical recognition record, denoted as original image quality indicators, and at the same time extract the image quality indicators after segmentation of each electrical component.

[0008] A segmentation duration extraction module is used to extract the segmentation duration of each electrical component from the historical recognition record.

[0009] A sensitive image quality indicator analysis module is used to compare the image quality indicators after segmentation of each electrical component in the historical recognition record with the original image quality indicators, thereby analyzing the sensitive image quality indicators of each electrical component.

[0010] A sensitive segmentation correlation analysis module is used to correlate the segmentation duration of the electrical component with the sensitive image quality indicators of the electrical component to obtain the segmentation duration sensitive correlation.

[0011] A current electrical component wiring extraction module is used to obtain the electrical components contained in the current electrical component wiring diagram, and thereby extract the sensitive image quality indicators of each electrical component.

[0012] A segmentation determination module is used to perform image quality detection on the current electrical component wiring diagram to obtain the original image quality indicators, and combine the sensitive image quality indicators of each electrical component and the segmentation duration sensitive correlation to determine the initial segmentation order of the electrical components and the image processing method.

[0013] A segmentation dynamic implementation module is used to perform real-time system resource detection during the segmentation of the current electrical component wiring diagram according to the initial segmentation order of the electrical components, and dynamically adjust the initial segmentation order of the electrical components according to the system resource detection results.

[0014] Compared with the prior art, the present application has the following advantages: 1. The present application extracts the original image quality indicators and the image quality indicators after segmentation of the electrical components by means of the historical recognition record of the electrical component wiring diagram, thereby identifying the sensitive image quality indicators of the electrical components by analyzing the fluctuations in the image quality indicators, providing a targeted processing goal for image processing during segmentation of each electrical component, effectively improving the purpose of image processing, providing image background conditions that meet actual needs for electrical component segmentation, and thereby maximizing the segmentation efficiency.

[0015] The application predicts the partition time by using the electrical elements in the current electrical element connection diagram, determines the initial partition order of the electrical elements according to the prediction result, and then detects the system resources in real time during the partition process according to the initial partition order, and dynamically adjusts the initial partition order according to the detection result. This method can realize adaptive adjustment based on the system resource limited condition, so that the electrical element partition is more in line with the actual situation of the system resources, thereby significantly reducing the occurrence rate of partition interruption and ensuring the continuity of the partition process and the integrity of the image data. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0017] Figure 1 The schematic diagram of the connection of the modules of the system of the present application.

[0018] Figure 2 The flowchart for analyzing the sensitive image quality indicators of each electrical element in the present application.

[0019] Figure 3 The implementation schematic diagram of whether to adjust the initial partition order of the electrical elements in the present application. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0021] The present application proposes an electrical element connection diagram recognition system using image partition technology, which includes a historical recognition record calling module, an image quality indicator extraction module, a partition time extraction module, a sensitive image quality indicator analysis module, a sensitive partition correlation analysis module, a current electrical element connection extraction module, a partition determination module and a partition dynamic implementation module.

[0022] Referring to Figure 1As shown, the historical recognition record calling module is connected with the image quality indicator extraction module and the segmentation duration extraction module, the image quality indicator extraction module is connected with the sensitive image quality indicator analysis module, the sensitive image quality indicator analysis module and the segmentation duration extraction module are connected with the sensitive segmentation association analysis module, the sensitive image quality indicator analysis module is connected with the current electrical element wiring extraction module, the current electrical element wiring extraction module and the sensitive segmentation association analysis module are connected with the segmentation determination module, and the segmentation determination module is connected with the segmentation dynamic implementation module.

[0023] It needs to be emphasized that the electrical element wiring diagram mentioned in the present application is generated by computer special drawing software. Due to the influence of factors such as display resolution of computer system, graphic rendering algorithm, file format conversion and output device, there may be certain differences in image quality of the generated electrical element wiring diagram.

[0024] Illustratively, the physical resolution of the computer display determines the sharpness of the image. Lower resolution may result in loss of details in the image, affecting the recognition accuracy of electrical elements and connection lines. In another illustrative example, the anti-aliasing algorithm used by the drawing software when generating the image may affect the smoothness of the lines. Different anti-aliasing algorithms may result in different sharpness of the line edges, thereby affecting the recognition effect of the electrical elements. In yet another illustrative example, lossy or lossless compression may be performed when saving the wiring diagram as different file formats. Lossy compression reduces image quality, while lossless compression preserves more details. The choice of file format directly affects the sharpness and recognizability of the image.

[0025] Therefore, the image quality of the electrical element wiring diagram generated by the computer is not high and stable, and image processing is needed in the electrical element recognition and segmentation.

[0026] The historical recognition record calling module is used to call the historical recognition record of the electrical element wiring diagram, and extract the electrical elements contained in the electrical element wiring diagram from the historical recognition record.

[0027] It needs to be known that in the identification of the electrical element wiring diagram, an identification record will be generated, which can be used as a reference for subsequent identification. The identification record usually contains information such as the original image of the electrical element wiring diagram, image quality detection results (such as resolution, contrast, signal-to-noise ratio, etc.), segmentation duration of each electrical element, image quality indicators after segmentation of each electrical element, etc. In addition, a special area (called "element list" or "symbol explanation area") is usually left in the electrical element wiring diagram for name labeling and explanation of the contained electrical elements, for easy reference and comparison.

[0028] The image quality indicator extraction module is configured to extract image quality indicators of the original electrical component wiring diagram from the historical identification records, denoted as original image quality indicators, and meanwhile extract image quality indicators of each segmented electrical component.

[0029] In the example of the above scheme, the image quality indicators mentioned above include but are not limited to resolution, contrast, signal-to-noise ratio, color depth, sharpness, etc.

[0030] Specifically, the resolution of an image refers to the number of pixels per unit length in the image. Generally, a high-resolution image contains more detailed information, which helps improve the recognition accuracy. A low-resolution image may result in loss of details, affecting the segmentation effect.

[0031] Contrast refers to the difference in brightness between the brightest and darkest parts of an image. Generally, the boundary between electrical components and background in a high-contrast image is more obvious, which helps improve the recognition accuracy of the segmentation algorithm. A low-contrast image may result in blurred boundaries, increasing the risk of false positives and false negatives.

[0032] Signal-to-noise ratio is the ratio of target information (i.e. useful signal) to background noise in an image. Generally, noise affects the accuracy of segmentation and recognition, especially for complex-shaped electrical components. Noise may change the boundaries of components, leading to misidentification or loss of shape features.

[0033] Color depth represents the number of colors that can be represented by each pixel in an image, specifically the number of bits per color channel (e.g. red, green, blue). Color depth determines the range and variety of colors that can be displayed in an image. High color depth can represent a wider variety of colors, providing more rich color information. For example, different colored wires may represent different voltage levels or signal types, and high color depth can ensure that these colors are accurately identified. Low color depth can only represent a limited variety of colors, which may result in color distortion or uneven transitions between colors, making it difficult to distinguish between certain color-similar electrical components and increasing the risk of recognition errors.

[0034] Sharpness refers to the clarity of edges in an image, reflecting the smoothness of transitions between different regions in the image. High sharpness images have clearer edges, which helps improve the recognition accuracy of the segmentation algorithm. Low sharpness images may have blurred edges, affecting the segmentation effect.

[0035] It needs to be explained that after completing the image quality detection of the original electrical component wiring diagram, it needs to be processed for image processing, such as denoising, enhancing contrast, sharpening, etc. to meet the needs of each electrical component identification and segmentation. Due to the differences in specific requirements of different electrical components for image quality, the performance of various image quality indicators in the segmented image will be different, which is mainly due to their physical characteristics, functional complexity, etc. The first specific embodiment is that the shapes and sizes of different electrical components are different, for example, resistors are usually simple cylindrical or rectangular, while integrated circuits may have complex pin arrangements and multi-layer structures. Complex shape and smaller size require higher resolution and edge clarity to ensure accurate identification and segmentation, so the image needs to have higher resolution and edge clarity when segmenting resistors.

[0036] Second: the colors and materials of electrical components are also different, some components (such as colored capacitors, light-emitting diodes LED) rely on specific colors for identification, while other components (such as relays with metal shells) may have reflective surfaces and are easily affected by lighting conditions. High color depth is essential for the identification of these components, so the image needs to have high color depth when segmenting colored capacitors.

[0037] Third: Some electrical components (such as high-frequency oscillators, radio frequency components) are very sensitive to noise, and any slight interference can affect their performance. Therefore, when identifying and segmenting these components, it is necessary to minimize noise in the image and maintain a high signal-to-noise ratio to ensure accuracy and stability. If the original electrical component wiring diagram does not have such image quality indicators, targeted image processing is needed when segmenting the corresponding electrical components to meet the segmentation requirements.

[0038] The segmentation duration extraction module is used to extract the segmentation duration of each electrical component from the historical identification record, and the specific operation is as follows: extract the electrical component, initial segmentation time and end segmentation time of each segmentation from the historical identification record.

[0039] Compare the end segmentation time of each segmented electrical component with the initial segmentation time to obtain the segmentation duration of each electrical component in the historical identification record.

[0040] The sensitive image quality indicator analysis module is used to compare the image quality indicators of each electrical component after segmentation in the historical identification record with the original image quality indicators, thereby analyzing the sensitive image quality indicators of each electrical component, see Figure 2As shown, the specific analysis is as follows: the image quality indicators of each electrical element after segmentation in each historical identification record are compared with the original image quality indicators to obtain image quality indicator comparison differences, and the absolute values of the image quality indicator comparison differences are divided by the original image quality indicators to obtain the fluctuation degrees of the image quality indicators of each electrical element after segmentation.

[0041] The electrical elements existing in all historical identification records are classified to constitute a plurality of historical identification records corresponding to each electrical element.

[0042] The fluctuation degrees of the image quality indicators of the same electrical element after segmentation in each historical identification record are compared, and the image quality indicator corresponding to the maximum fluctuation degree is selected as the sensitive image quality indicator of the same electrical element in each historical identification record.

[0043] Again, it needs to be explained that if a certain image quality indicator significantly fluctuates after segmentation of a certain electrical element, it indicates that more intensive processing is performed on this indicator during segmentation of this electrical element. This phenomenon indirectly reflects that the electrical element has a higher dependence and sensitivity to this image quality indicator, which means that fine adjustment of this indicator must be made during segmentation to ensure the accuracy of identification and segmentation. Applied to the above explanation, the image quality indicator with smaller fluctuation degree usually means that the identification and segmentation process of the electrical element is more stable and less affected by external factors, and the image quality indicator with larger fluctuation degree usually means that more processing is performed on this indicator during segmentation of the electrical element. For example, if the fluctuation degree of color depth is large, it means that the color depth is significantly adjusted in some identification records to improve the identification accuracy, and the color depth can be used as the sensitive image quality indicator.

[0044] The sensitive image quality indicators of the same electrical element in each historical identification record are compared, the mode is extracted, the occurrence frequency of each mode is obtained, the mean and standard deviation of the occurrence frequency of each mode are calculated for discrete coefficient statistics, and then the discrete coefficient is compared with the set threshold value. If the discrete coefficient reaches the set threshold value, the maximum mode is taken as the sensitive image quality indicator corresponding to the corresponding electrical element, otherwise the mode with occurrence frequency higher than the median occurrence frequency is taken as the sensitive image quality indicator corresponding to the corresponding electrical element.

[0045] It needs to be understood that the mode is considered when the sensitive image quality indicators of the same electrical element in each historical identification record are used to determine the sensitive image quality indicators of the electrical element because the mode refers to the value that frequently appears in a set of data, and by extracting the mode of the sensitive image quality indicators of the same electrical element in each historical identification record, those indicators that frequently appear in multiple identification processes can be found, and these indicators usually have a greater impact on the identification of the electrical element. When there is more than one mode, the dispersion coefficient of each mode is calculated, and the dispersion coefficient is the ratio of the standard deviation to the mean for measuring the relative dispersion degree of the data. The greater the dispersion coefficient, the greater the volatility of the data, and the smaller the dispersion coefficient, the more stable the data. By comparing the dispersion coefficient with the preset threshold value, when the dispersion coefficient reaches or exceeds the threshold value, it indicates that the frequency distribution of the mode is relatively dispersed, and there are multiple important image quality indicators. At this time, the largest mode is selected as the sensitive image quality indicator to ensure that the most frequently occurring and most influential indicator is selected. Conversely, when the dispersion coefficient is less than the threshold value, it indicates that the frequency distribution of the mode is relatively concentrated, and at this time, the mode with a frequency higher than the median frequency is selected as the sensitive image quality indicator to ensure that the relatively stable and more influential indicator for identification is selected.

[0046] When setting the threshold value of the dispersion coefficient, it is considered that the dispersion coefficient is the ratio of the standard deviation to the mean, and when the standard deviation exceeds the mean, the dispersion coefficient will be greater than 1, which indicates that the difference between the data points has significantly increased, and the frequency distribution has become extremely dispersed. Therefore, the dispersion coefficient equal to 1 can be used as a critical point because it marks the turning point of the data from relative stability to high volatility. When the dispersion coefficient is less than 1, the data is relatively concentrated and less volatile, and when the dispersion coefficient is greater than 1, the volatility of the data significantly increases, and the frequency distribution becomes more dispersed. In this case, 0.8 is selected as the threshold value to find a reasonable balance between stability and flexibility. 0.8 is slightly lower than the critical point 1, allowing for some volatility, but not too loose, thereby ensuring the robustness and accuracy of the identification system.

[0047] Based on the basic principles of statistics, the mode, frequency, mean, standard deviation, and dispersion coefficient are used to objectively evaluate the importance of different image quality indicators. This method not only considers the absolute value of the frequency, but also considers the relative distribution of the frequency, avoiding the limitations of a single indicator.

[0048] The sensitive segmentation association analysis module is used to associate the segmentation duration of the electrical element with the sensitive image quality indicators of the electrical element to obtain the segmentation duration sensitive association. The specific implementation process is as follows: based on the sensitive image quality indicators corresponding to each electrical element, the data of the sensitive image quality indicators in the original image quality indicators is extracted from the historical identification records corresponding to each electrical element, which is recorded as the original sensitive image quality indicators.

[0049] The several historical identification records corresponding to each electrical element are formed into the split time length sensitive correlation curve corresponding to each electrical element in the coordinate system with the original sensitive image quality index as the horizontal axis and the split time length as the vertical axis.

[0050] It should be added that in order to improve the accuracy of identification and segmentation during the electrical element segmentation process, the original image usually needs to be preprocessed, especially the processing of the sensitive image quality index. The complexity and computational amount of these processing steps directly affect the split time length, so the split sensitive correlation curve can be constructed by means of the original sensitive image quality index and the split time length corresponding to each historical identification record in the electrical element segmentation to reflect the correlation between the split time length and the sensitive image quality index.

[0051] The current electrical element wiring extraction module is used to obtain the electrical elements contained in the current electrical element wiring diagram, which can be obtained from the element list of the wiring diagram, and then match the sensitive image quality index of each electrical element obtained from the historical identification record with the sensitive image quality index of each electrical element contained in the current electrical element wiring diagram, so as to obtain the sensitive image quality index of each electrical element in the current electrical element wiring diagram.

[0052] The split determination module is used to detect the original image quality characteristics of the current electrical element wiring diagram, and determine the initial split order of the electrical elements and the image processing method in combination with the sensitive image quality index of each electrical element and the split time length sensitive correlation.

[0053] In the preferred operation of the above scheme, the initial split order of the electrical elements is determined by referring to the following process: based on the sensitive image quality index of each electrical element in the current electrical element wiring diagram, the split time length sensitive correlation curve corresponding to each electrical element in the current electrical element wiring diagram is retrieved from the split time length sensitive correlation curve corresponding to each electrical element.

[0054] According to the original image quality index of the current electrical element wiring diagram and the split time length sensitive correlation curve corresponding to each electrical element, the split time length of each electrical element in the current electrical element wiring diagram is predicted, and the prediction is as follows: the number of sensitive image quality indexes corresponding to each electrical element in the current electrical element wiring diagram is counted, if the sensitive image quality index corresponding to an electrical element has only one, the split time length corresponding to the original sensitive image quality index is extracted from the split time length sensitive correlation curve corresponding to the electrical element using the original sensitive image quality index in the current electrical element wiring diagram as the predicted split time length corresponding to the electrical element.

[0055] If there are more than one sensitive image quality indicators corresponding to an electrical component, the original sensitive image quality indicators corresponding to the electrical component are extracted from the sensitive duration association curve of the electrical component, and the original sensitive image quality indicators corresponding to the electrical component are extracted from the sensitive duration association curve of the electrical component.

[0056] The range of the predicted split duration of the electrical component on each sensitive image quality indicator is calculated, and the calculation result is compared with the set critical value, for example, the critical value is 0.3, if the range does not reach the critical value, it means that the influence of different sensitive image quality indicators on the split duration is consistent, and the mean value of the split duration can be taken as the final predicted split duration; if the range reaches or exceeds the critical value, it means that the influence of different sensitive image quality indicators on the split duration is different, at this time, the weight value of each sensitive image quality indicator corresponding to the electrical component is assigned, and then the predicted split duration of the electrical component on each sensitive image quality indicator is combined with the weight value to calculate the predicted split duration of the electrical component. Weighted average can better reflect the comprehensive influence of different sensitive image quality indicators on the split duration, and ensure the accuracy of the prediction result.

[0057] It should be noted that the weight value of each sensitive image quality indicator in the above can be the frequency of occurrence of each sensitive image quality indicator as the weight value.

[0058] The present application can effectively distinguish whether the influence of different sensitive image quality indicators on the split duration is consistent by introducing range calculation and critical value comparison. If the influence is consistent, the mean value can be taken; if the influence is different, the contribution of each sensitive image quality indicator can be more accurately reflected through weight assignment and weighted average, and the prediction accuracy is improved.

[0059] The electrical components are arranged in order of split duration from long to short to obtain an initial split order of the electrical components.

[0060] It should be noted that the electrical components with longer split duration are preferentially split because the electrical components with longer split duration usually mean that these components have higher requirements for image quality, and preferentially processing these complex components can ensure that they are fully processed in the early stage, thereby simplifying the subsequent split task.

[0061] In the further preferred operation of the above scheme, the image processing mode is determined as follows: the sensitive image quality indicators of the electrical components are extracted in order according to the initial split order of the electrical components, and the corresponding original sensitive image quality indicators of the current electrical component connection diagram are preferentially processed before the split operation of the electrical components, which can improve the pertinence of image processing and significantly improve the accuracy and efficiency of electrical component splitting.

[0062] The segmentation implementation module is used for system resource detection in the segmentation implementation of the current electrical element wiring diagram according to the initial electrical element segmentation sequence. Specifically, the system resource detection is real-time system resource indicator parameter detection during the identification and segmentation process of the current electrical element wiring diagram. The system resource indicators include CPU usage, memory usage, and disk usage. The initial electrical element segmentation sequence is dynamically adjusted according to the detection results.

[0063] The reasons for selecting CPU usage, memory usage, and disk usage as system resource indicators in the above are as follows: First, the identification and segmentation of electrical elements usually involve a large number of computing operations, such as image processing and feature extraction. These operations require high computing power of the CPU. By monitoring the CPU usage in real time, the situation of excessive CPU load can be found in time. Second, the electrical element wiring diagram usually contains a large amount of image data, especially high-resolution images or multiple images. These data need to occupy a large amount of memory space for storage and processing. By monitoring the memory usage in real time, it can be ensured that the system has enough memory to support the identification and segmentation task. Finally, the identification and segmentation task of the electrical element wiring diagram usually needs to load image data from the disk. If the disk reading speed is slow, it may cause delay in the identification and segmentation process, affecting the overall efficiency. By monitoring the disk usage in real time, it can be ensured that the disk has enough bandwidth to support fast data reading and writing operations.

[0064] In the innovative implementation of the above scheme, the initial electrical element segmentation sequence is dynamically adjusted according to the detection results as follows: whether there is resource limitation is identified according to the system resource detection results. Specifically, the detected system resource indicators are compared with the corresponding warning thresholds. If a certain system resource indicator reaches the warning threshold, it is identified that there is resource limitation. If all system resource indicators do not reach the warning threshold, it is identified that there is no resource limitation.

[0065] The warning thresholds of the system resource indicators in the above can be obtained from the system usage instructions according to the system model. This is because different models of hardware devices and system configurations have significant differences in performance, heat dissipation, power consumption, etc. Manufacturers usually provide recommended warning thresholds in the usage instructions based on the design and test results of their products to ensure that the system runs in the best state. Exemplarily, the recommended warning thresholds are as follows: CPU usage: the warning threshold is 80%. When the CPU usage continuously exceeds this threshold, it indicates that the system may face a situation of tight computing resources.

[0066] Memory usage: the warning threshold is 80%. When the memory usage continuously exceeds this threshold, there may be a risk of insufficient memory.

[0067] Disk usage rate: The warning threshold should be set according to the specific disk type and application scenario. For mechanical hard disks, the recommended warning threshold is 70% to avoid I / O bottlenecks; for solid-state disks, the recommended warning threshold is 85% because SSDs have faster read and write speeds and can tolerate higher usage rates.

[0068] Referring to Figure 3 As shown, if there is no resource limitation at a certain detection time, the initial electrical element segmentation order is not adjusted, and if there is resource limitation at a certain detection time, the electrical element to be segmented is extracted from the initial electrical element segmentation order as the target electrical element, and the electrical element at the end of the initial electrical element segmentation order is extracted as the adjusted target electrical element.

[0069] The adjusted target electrical element is moved to the position to be segmented, and the segmentation order is regenerated.

[0070] It needs to be explained that when the resource limitation is identified, the electrical element at the end is extracted to the position to be segmented because the electrical element at the end has a shorter segmentation time, which can ensure that the electrical element with shorter segmentation time has the opportunity to be processed preferentially when the resource is limited, which helps to evenly distribute the processing load. In addition, when the resource is limited, only one end electrical element is adjusted each time, which can avoid frequent order adjustment and maintain the relative stability of the segmentation order.

[0071] The above content is only an example and explanation of the concept of the present application. Those skilled in the art can make various modifications or supplements to the described specific embodiments or use similar ways to replace them, as long as they do not deviate from the concept of the present application or exceed the scope defined by the present application.

Claims

1. An electrical component connection diagram recognition system using an image segmentation technique, characterized by, The method comprises the following modules: a historical identification record calling module, configured to call historical identification records of an electrical element wiring diagram and extract electrical elements contained in the electrical element wiring diagram from the historical identification records; an image quality indication extraction module, configured to extract image quality indications of an original electrical element wiring diagram from the historical identification records, denoted as original image quality indications, and extract image quality indications after segmentation of each electrical element at the same time; a segmentation duration extraction module, configured to extract segmentation durations of each electrical element from the historical identification records; a sensitive image quality indication analysis module, configured to compare the image quality indications after segmentation of each electrical element in the historical identification records with the original image quality indications, so as to analyze sensitive image quality indications of each electrical element; a sensitive segmentation correlation analysis module, configured to correlate the segmentation durations of the electrical elements with the sensitive image quality indications of the electrical elements to obtain segmentation duration sensitive correlations; a current electrical element wiring extraction module, configured to obtain electrical elements contained in a current electrical element wiring diagram and extract sensitive image quality indications of each electrical element according to the electrical elements; a segmentation determination module, configured to perform image quality detection on the current electrical element wiring diagram to obtain original image quality indications, and determine an initial segmentation order of the electrical elements and an image processing mode by combining the original image quality indications, the sensitive image quality indications of each electrical element and the segmentation duration sensitive correlations; a segmentation dynamic implementation module, configured to perform real-time system resource detection during segmentation of the current electrical element wiring diagram according to the initial segmentation order of the electrical elements, and dynamically adjust the initial segmentation order of the electrical elements according to the system resource detection results.

2. The electrical component connection diagram recognition system using an image segmentation technique according to claim 1, characterized by: The extraction of the segmentation durations of each electrical element from the historical identification records is implemented as follows: extracting electrical elements, initial segmentation times and end segmentation times of each segmentation from the historical identification records; comparing the end segmentation times with the initial segmentation times of each segmented electrical element to obtain the segmentation durations of each electrical element in the historical identification records.

3. The electrical component connection diagram recognition system using an image segmentation technique according to claim 1, wherein: The analysis of the sensitive image quality indications of each electrical element is implemented as follows: comparing the image quality indications after segmentation of each electrical element in each historical identification record with the original image quality indications to obtain image quality indication comparison differences, taking absolute values of the image quality indication comparison differences and dividing the absolute values by the original image quality indications to obtain image quality indication fluctuation degrees after segmentation of each electrical element; classifying electrical elements existing in all historical identification records to form a plurality of historical identification records corresponding to each electrical element; comparing the image quality indication fluctuation degrees after segmentation of the same electrical element in each historical identification record, and selecting an image quality indication corresponding to a maximum fluctuation degree as a sensitive image quality indication of the same electrical element in each historical identification record; and a segmentation dynamic implementation module, configured to perform real-time system resource detection during segmentation of the current electrical element wiring diagram according to the initial segmentation order of the electrical elements, and dynamically adjust the initial segmentation order of the electrical elements according to the system resource detection results. The sensitive image quality indicators of the same electrical element in each historical identification record are compared, the mode is extracted, the frequency of each mode is obtained, the mean and standard deviation of the frequency of each mode are calculated to obtain the dispersion coefficient, and then the dispersion coefficient is compared with the set threshold value. If the dispersion coefficient reaches the set threshold value, the maximum mode is taken as the sensitive image quality indicator corresponding to the corresponding electrical element, otherwise the mode with a frequency higher than the median frequency is taken as the sensitive image quality indicator corresponding to the corresponding electrical element.

4. The electrical component connection diagram recognition system using an image segmentation technique according to claim 1, wherein: The split duration sensitive correlation is analyzed as follows: Based on the sensitive image quality indicators corresponding to each electrical element, the data of the sensitive image quality indicators in the original image quality indicators is extracted from the several historical identification records corresponding to each electrical element, which is recorded as the original sensitive image quality indicator; The several historical identification records corresponding to each electrical element are arranged in a coordinate system with the original sensitive image quality indicator as the horizontal axis and the split duration as the vertical axis to form a split duration sensitive correlation curve corresponding to each electrical element.

5. The electrical component connection diagram recognition system using an image segmentation technique according to claim 4, wherein: The initial split order of the electrical element is determined as follows: According to the original image quality indicator of the current electrical element wiring diagram and the split duration sensitive correlation curve corresponding to each electrical element, the split duration of each electrical element in the current electrical element wiring diagram is predicted; The electrical elements are arranged in order from long to short according to the split duration to obtain the initial split order of the electrical elements.

6. The electrical component connection diagram recognition system using an image segmentation technique according to claim 5, wherein: The split duration of each electrical element in the current electrical element wiring diagram is predicted as follows: The number of sensitive image quality indicators corresponding to each electrical element in the current electrical element wiring diagram is counted. If the sensitive image quality indicator corresponding to an electrical element is only one, the original sensitive image quality indicator in the current electrical element wiring diagram is used to extract the split duration corresponding to the original sensitive image quality indicator from the split duration sensitive correlation curve corresponding to the electrical element as the predicted split duration corresponding to the electrical element. If the sensitive image quality indicator corresponding to an electrical element is more than one, the split duration corresponding to each original sensitive image quality indicator is extracted from the split duration sensitive correlation curve corresponding to the electrical element using the original sensitive image quality indicator in the current electrical element wiring diagram as the predicted split duration of the electrical element on each sensitive image quality indicator. The range of the predicted split duration of the electrical element on each sensitive image quality indicator is calculated, and the calculation result is compared with the set critical value. If the critical value is not reached, the mean of the predicted split duration of the electrical element on each sensitive image quality indicator is taken as the predicted split duration corresponding to the electrical element. If the critical value is reached, each sensitive image quality indicator corresponding to the electrical element is weighted, and then the predicted split duration of the electrical element on each sensitive image quality indicator is combined with the weight value to obtain the predicted split duration corresponding to the electrical element.

7. The electrical component connection diagram recognition system using an image segmentation technique according to claim 1, wherein: The image processing method is determined as follows: According to the initial segmentation order of the electrical components, sensitive image quality indicators of the electrical components are extracted in sequence, and then the corresponding original sensitive image quality indicators of the current electrical component wiring diagram are preferentially processed before the segmentation operation of the electrical components.

8. The electrical component connection diagram recognition system using an image segmentation technique according to claim 1, wherein: The system resource detection is as follows: The system resource indication parameters are detected in real time during the identification and segmentation process of the current electrical component wiring diagram, wherein the system resource indication parameters include CPU usage, memory usage and disk usage.

9. The electrical component connection diagram recognition system using an image segmentation technique according to claim 8, wherein: The initial segmentation order of the electrical components is dynamically adjusted according to the system resource detection results as follows: According to the system resource detection results, it is identified whether there is resource limitation. If there is no resource limitation at a certain detection time, the initial segmentation order of the electrical components is not adjusted. If there is resource limitation at a certain detection time, the electrical component to be segmented from the initial segmentation order of the electrical components is extracted as the target electrical component, and the electrical component at the end of the initial segmentation order of the electrical components is extracted as the adjusted target electrical component. The adjusted target electrical component is moved to the position to be segmented, and the segmentation order is regenerated.

10. The electrical component connection diagram recognition system using an image segmentation technique according to claim 9, wherein: Whether there is resource limitation is identified as follows: The detected system resource indication parameters are compared with the corresponding indication parameter warning thresholds respectively. If a certain system resource indication parameter reaches the warning threshold, it is identified that there is resource limitation. If all system resource indication parameters do not reach the warning threshold, it is identified that there is no resource limitation.

Images