Identification analysis system based on artificial intelligence

Abstract

The invention discloses a recognition and analysis system based on artificial intelligence. An operation method of the system comprises the following steps: step 1, carrying out image acquisition and labeling on road cracks; 2, performing road crack recognition model training; 3, evaluating and analyzing the detected cracks; step 4, crack detection result display and evaluation: an image acquisition and labeling module is used for acquiring road crack images and performing crack labeling; the crack detection and identification module is used for establishing a detection model to realize identification of road cracks; the crack analysis and evaluation module is used for analyzing the detected cracks and evaluating the service life of the pavement; the crack labeling module is used for carrying out manual labeling on the preprocessed image; the crack detection module is used for automatically identifying and positioning a crack area in the image; the method has the characteristics of timely road crack detection and accurate crack identification.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and specifically to an identification and analysis system based on artificial intelligence. Background Art

[0002] In modern society, road transportation plays an important role, and the crack problem on the road surface is an important factor affecting traffic safety and road sustainability. The existence of cracks on the road surface will not only increase the driving risk of drivers, leading to traffic accidents, but also have an adverse impact on the controllability and driving comfort of vehicles. In addition, the cracks will further expand and intensify, resulting in more serious road damage, increasing maintenance and repair costs, and reducing the service life of the road.

[0003] At present, traditional road crack detection and evaluation methods mainly rely on manual inspections, which have the following deficiencies: Subjectivity: Manual inspections rely on the subjective judgment and experience of personnel, and the results may be affected by factors such as individual differences, fatigue, and environmental conditions, resulting in problems of misjudgment and inconsistency. Inefficiency: Manual inspections require a large amount of manpower and time investment, especially for large-scale road networks, which are often time-consuming and laborious, and cannot achieve real-time monitoring. Lack of quantitative evaluation: The results of manual inspections usually can only provide the approximate location and degree of cracks, lacking accurate measurement and evaluation of crack parameters (such as length, width, depth, etc.). Therefore, it is necessary to design an identification and analysis system based on artificial intelligence with good timeliness and accuracy for crack detection. Summary of the Invention

[0004] The purpose of the present invention is to provide an identification and analysis system based on artificial intelligence to solve the problems raised in the above background art.

[0005] To solve the above technical problems, the present invention provides the following technical solution: An identification and analysis system based on artificial intelligence, and the operation method of the system includes the following steps: Step 1: Image acquisition and annotation of road cracks; Step 2: Training of the road crack recognition model; Step 3: Evaluation and analysis of the detected cracks; Step 4: Display and evaluation of the crack detection results.

[0006] According to the above technical solution, the step of image acquisition and annotation of road cracks includes: Select a suitable image acquisition device to ensure that high-quality road crack images can be obtained; Perform image preprocessing, including denoising, enhancement, and size calibration; Through manual annotation, label the parameter information of the cracks on the image.

[0007] According to the above technical solution, the step of manually annotating the parameter information of the crack on the image includes: On the collected road crack image, the crack is annotated by professionals or image annotation algorithms. The main contents of the annotation include the width, depth, length, and density information of the crack. Annotation tools or automated algorithms can be used to represent the position and parameters of the crack by drawing polygons, line segments, or pixel-level marks. Finally, the collected and annotated road crack images are organized into a data set, including a training set and a validation set. To improve the generalization ability of the model, it is necessary to ensure that the data set contains diverse road crack samples, covering cracks of different types, degrees, and shapes. Through the above steps, a road crack image data set containing annotated crack parameters can be established, providing valuable data resources for the training and validation of subsequent crack detection and analysis models.

[0008] According to the above technical solution, the steps of training the road crack recognition model include: Using the convolutional neural network algorithm to train the model with crack data; Using an independent test set to test and validate the trained model.

[0009] According to the above technical solution, the steps of evaluating and analyzing the detected crack include: Measuring the length, width, depth, and density parameters of the crack; Setting a preset parameter threshold to determine whether the crack parameters exceed the threshold; Recording the crack image information for subsequent use.

[0010] According to the above technical solution, the steps of measuring the length, width, depth, and density parameters of the crack include: Measuring the length, width, depth, and density parameters of each detected crack. Using computer vision and image processing techniques, through scale calibration and pixel calculation in the image, the actual size of the crack is obtained. For example, based on the known scale information in the image, the pixel length of the crack is calculated, and the actual length of the crack is obtained through the proportional relationship with the actual length. For the depth of the crack, a non-contact imaging measurement method is adopted: using a high-resolution camera or laser imaging system to image the road crack, and then using image processing and computer vision techniques to analyze the brightness or texture changes in the image to estimate the depth of the crack. The crack density refers to the number of cracks per unit area. By counting the cracks in the image, the density information of the crack can be obtained. Image segmentation and image processing techniques can be used to separate the crack area from the background and analyze and count the cracks. The calculation of the crack density can provide the overall distribution of road surface cracks.

[0011] According to the above technical solution, the step of recording crack image information for subsequent use includes: Record and store the crack parameters obtained through measurement and statistics for subsequent data analysis and maintenance plan formulation. At the same time, establish a database or file system to store the parameter information of each detected crack, as well as the corresponding image data and timestamps. This can achieve long-term monitoring and analysis of road crack conditions, provide a decision-making basis for road maintenance and repair. Through the above steps, detailed parameter measurement and statistical analysis of the detected cracks can be carried out, the severity of the cracks can be judged according to the preset threshold, and automated crack information reporting and repair arrangements can be realized, thereby improving the efficiency and accuracy of road maintenance.

[0012] According to the above technical solution, the step of crack detection result display and evaluation includes: The results of crack detection and analysis are presented to the administrator in a visual form; Carry out the evaluation and prediction of road service life and maintenance requirements.

[0013] According to the above technical solution, the step of carrying out the evaluation and prediction of road service life and maintenance requirements includes: According to the type and degree of the cracks, the system can evaluate and predict the road service life. Based on the existing road materials, design standards and historical data, through the comprehensive analysis of crack parameters and other relevant factors, the system estimates the remaining service life of the road and provides corresponding suggestions and maintenance requirements. This will help the administrator formulate a reasonable road maintenance plan and budget to extend the road service life and ensure traffic safety.

[0014] According to the above technical solution, the system includes: An image acquisition and annotation module for acquiring road crack images and performing crack annotation; A crack detection and recognition module for establishing a detection model to realize the recognition of road cracks; A crack analysis and evaluation module for analyzing the detected cracks and evaluating the road surface life.

[0015] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: The present invention realizes the automated detection, analysis and evaluation of road cracks through the three modules of image acquisition and annotation, crack acquisition and recognition, and crack analysis and evaluation. By using deep learning algorithms for crack detection and recognition, combined with parameter measurement and statistical analysis, the system can provide the ability to quantitatively evaluate road cracks, predict road service life and maintenance requirements, and provide decision-making support for road maintenance and planning. The implementation of this patent can improve road maintenance efficiency, reduce the cost of manual inspection, and enhance road safety and sustainability. Brief Description of the Drawings

[0016] The drawings are used to provide a further understanding of the present invention and form a part of the specification. They are used together with the embodiments of the present invention to explain the present invention and do not constitute a limitation to the present invention. In the drawings: Figure 1 is a flowchart of an identification and analysis method based on artificial intelligence provided in the first embodiment of the present invention; Figure 2 is a schematic diagram of the module composition of an identification and analysis system based on artificial intelligence provided in the second embodiment of the present invention. Detailed Embodiments

[0017] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0018] Embodiment 1: Figure 1 is a flowchart of an identification and analysis method based on artificial intelligence provided in the first embodiment of the present invention. This embodiment can be applied to the scenario of road crack detection and analysis. This method can be executed by an identification and analysis system based on artificial intelligence provided in this embodiment, as Figure 1 shown. The method specifically includes the following steps: Step 1: Collect and annotate images of road cracks; In the embodiment of the present invention, images of road cracks are collected and annotated to establish a data set required for training and validating the model; Exemplarily, a high-resolution digital camera or a dedicated road monitoring camera is selected as the image collection device. These devices can capture the crack details on the road surface and provide clear images. Subsequently, different types of roads are selected, such as highways and urban roads, as well as roads in different geographical locations and sections known to have crack problems, to increase the width and breadth of the samples. During the image collection process, appropriate exposure time, aperture size, focal length parameters, etc. are set according to different weather conditions to ensure the quality of the collected images; Exemplarily, first, the collected images are preprocessed to improve the image quality. Subsequently, the collected images are labeled. On the collected road crack images, professionals or image annotation algorithms are used to annotate the cracks. The annotation content mainly includes information such as the width, depth, length, and density of the cracks. Annotation tools or automated algorithms can be used to represent the position and parameters of the cracks by drawing polygons, line segments, or pixel-level marks. Finally, the collected and annotated road crack images are organized into a dataset, including a training set and a validation set. To improve the generalization ability of the model, it should be ensured that the dataset contains diverse road crack samples, covering different types, degrees, and shapes of cracks. Through the above steps, a road crack image dataset containing annotated crack parameters can be established, providing valuable data resources for the training and validation of subsequent crack detection and analysis models.

[0019] Step 2: Train the road crack recognition model; In the embodiment of the present invention, using the preprocessed image data and the corresponding crack annotation data, the model is trained through the convolutional neural network algorithm. The training objective is to enable the model to automatically identify and locate the crack regions in the images; Exemplarily, when training the crack detection model, first, a dataset of road cracks needs to be prepared and divided into a training set, a validation set, and a test set to ensure that the dataset contains road crack samples of different types, degrees, and shapes. Then, data augmentation operations are performed on the images in the training set, such as flipping, rotating, scaling, and adding noise, etc., to increase the diversity of the data and the robustness of the model. Next, a suitable crack detection model, such as the ResNet convolutional neural network, is selected and configured, and parameters such as the network layer structure, activation function, and optimization algorithm are set. The objective function of the crack detection task is defined. Usually, binary cross-entropy is used as the loss function to measure the difference between the model prediction result and the true label; Exemplarily, when training the model, the images in the training set are used as the input, and the weights and biases of the model are adjusted through the backpropagation algorithm and the optimizer, enabling the model to accurately predict the crack regions in the images. Through the iterative training and validation process, the hyperparameters of the model, such as the learning rate, batch size, regularization coefficient, etc., are tuned to optimize the performance of the model. During the training process, the training loss and validation loss are monitored to ensure that the model does not overfit or underfit. The accuracy, recall, precision, etc. of the model are evaluated using the validation set, and the model is further adjusted and optimized to improve its performance; Exemplarily, an independent test set is used to test and validate the trained model to evaluate the overall performance and generalization ability of the model. Through such a training process, the crack detection model can learn and extract the features of road cracks, realize the automatic identification and location of crack areas in images. Such a training process enables the model to have high accuracy and robustness, and achieve efficient road crack identification and analysis. Through the above steps, an electronic camera can be used to collect road surface images at regular intervals, and deep learning algorithms such as convolutional neural networks can be used to detect cracks in the preprocessed images. This artificial intelligence-based automated crack detection method can improve efficiency, reduce labor costs, and provide timely data support for subsequent crack analysis and repair work.

[0020] Step 3: Evaluate and analyze the detected cracks; In the embodiment of the present invention, parameter measurement and statistical analysis are performed on the cracks detected by the crack detection model. This step aims to obtain more detailed crack information and judge the severity of the cracks according to a preset threshold; Exemplarily, parameters such as length, width, and depth of each detected crack are measured. Using computer vision and image processing techniques, through scale calibration and pixel calculation in the image, the actual size of the crack is obtained. For example, based on the known scale information in the image, the pixel length of the crack is calculated, and the actual length of the crack is obtained through the proportional relationship with the actual length. For the depth of the crack, a non-contact imaging measurement method is adopted: a high-resolution camera or a laser imaging system is used to image the road crack, and then image processing and computer vision techniques are used to analyze the brightness or texture changes in the image to estimate the depth of the crack. The crack density refers to the number of cracks per unit area. By counting the cracks in the image, the density information of the cracks can be obtained. Image segmentation and image processing techniques can be used to separate the crack area from the background and analyze and count the cracks. The calculation of the crack density can provide the overall distribution of road surface cracks; Exemplarily, for crack parameters, a threshold is preset to determine the severity of the crack. When the length, width, depth, or density of the crack exceeds the preset threshold, the system will automatically report the crack information to the management personnel. The reported information can include the location, size, severity, etc. of the crack, so that the management personnel can take repair and restoration measures in a timely manner, record and store the crack parameters obtained through measurement and statistics for subsequent data analysis and maintenance plan formulation. At the same time, a database or file system is established to store the parameter information of each detected crack, as well as the corresponding image data and timestamp. In this way, long-term monitoring and analysis of the road crack condition can be achieved, providing a decision-making basis for road maintenance and repair. Through the above steps, detailed parameter measurement and statistical analysis of the detected cracks can be carried out, the severity of the cracks can be judged according to the preset threshold, and automated crack information reporting and repair arrangements can be realized, thereby improving the efficiency and accuracy of road maintenance.

[0021] Step Four: Display and Evaluation of Crack Detection Results.

[0022] In the embodiment of the present invention, the results of crack detection and analysis are presented to the administrator in a visual manner. At the same time, based on the type and degree of the crack, the system evaluates and predicts the service life and maintenance requirements of the road; Exemplarily, by visualizing the results of crack detection and analysis, the administrator can intuitively understand the distribution and severity of road cracks. This can be achieved by annotating the cracks on the original road image and drawing the crack areas. Different colors or marks can be used for annotation to represent the type and degree of the cracks, so that the administrator can quickly identify and understand the crack situation. In addition to image annotation, the system will also generate a crack parameter report, which includes detailed information on parameters such as the length, width, depth, and density of the cracks. The report is presented in the form of tables, charts, or text, clearly listing the parameter values and statistical data of each crack. This will provide the administrator with a comprehensive understanding of the cracks and support subsequent decision-making and planning work; Exemplarily, based on the type and degree of the crack, the system can evaluate and predict the service life of the road. Based on existing road materials, design standards, historical data, etc., through comprehensive analysis of crack parameters and other relevant factors, the system estimates the remaining service life of the road and provides corresponding suggestions and maintenance requirements. This will help the administrator formulate a reasonable road maintenance plan and budget to extend the service life of the road and ensure traffic safety. Through the above steps, the results of crack detection and analysis can be presented in an intuitive and clear manner, providing the administrator with a comprehensive road condition assessment and maintenance decision-making support. Such a result display and evaluation process will help optimize the road maintenance strategy and improve the reliability and sustainability of the road.

[0023] Example 2: Example 2 of the present invention provides an identification and analysis system based on artificial intelligence. Figure 2 It is a schematic diagram of the module composition of an identification and analysis system based on artificial intelligence provided in Example 2 of the present invention. As Figure 2 shown, the system includes: An image acquisition and annotation module, which is used to acquire road crack images and perform crack annotation; A crack detection and identification module, which is used to establish a detection model to realize the identification of road cracks; A crack analysis and evaluation module, which is used to analyze the detected cracks and evaluate the pavement life; In some embodiments of the present invention, the image acquisition and annotation module includes: A road image acquisition module, which is used to periodically acquire images of the road surface; A crack image preprocessing module, which is used to preprocess the acquired road images, including image denoising, enhancement, size calibration, etc.; A crack annotation module, which is used to manually annotate the preprocessed images, and annotate information such as the width, depth, length, and density of the cracks; In some embodiments of the present invention, the crack detection and identification module includes: An image feature extraction module, which is used to extract crack features from the preprocessed images through deep learning algorithms such as convolutional neural networks; A crack detection module, which is used to automatically identify and locate the crack areas in the images; A crack classification module, which is used to classify the detected cracks, and mark them according to the type and degree of the cracks for subsequent analysis and evaluation; In some embodiments of the present invention, the crack analysis and evaluation module includes: A crack parameter measurement module, which is used to measure the parameters of the detected cracks, including length, width, depth, and calculate the density of the cracks, etc.; A crack statistical analysis module, which is used to analyze the cracks based on crack parameters and statistical methods; A prediction and evaluation module, which is used to predict the service life and maintenance requirements of the road, and provide decision support and planning reference.

[0024] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0025] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An identification and analysis method based on artificial intelligence, characterized in that: The method includes the following steps: Step 1: Image acquisition and annotation of road cracks; Step 2: Training of the road crack recognition model; Step 3: Evaluation and analysis of the detected cracks; Step 4: Display and evaluation of the crack detection results.

2. The identification and analysis method based on artificial intelligence according to claim 1, characterized in that: The step of image acquisition and annotation of road cracks includes: Selecting a suitable image acquisition device to ensure high-quality road crack images can be obtained; Performing image preprocessing, including denoising, enhancement, and size calibration; Through manual annotation, the parameter information of the cracks is annotated on the image.

3. The identification and analysis method based on artificial intelligence according to claim 2, wherein: The step of annotating the parameter information of the cracks on the image through manual annotation includes: On the collected road crack images, cracks are annotated by professionals or image annotation algorithms. The main content of the annotation includes the width, depth, length, and density information of the cracks. Annotation tools or automated algorithms can be used to represent the position and parameters of the cracks by drawing polygons, line segments, or pixel-level marks. Finally, the collected and annotated road crack images are organized into a dataset, including a training set and a validation set. To improve the generalization ability of the model, it should be ensured that the dataset contains diverse road crack samples, covering different types, degrees, and shapes of cracks. Through the above steps, a road crack image dataset containing annotated crack parameters can be established, providing valuable data resources for the training and validation of subsequent crack detection and analysis models.

4. An identification and analysis method based on artificial intelligence according to claim 1, characterized in that: The step of training the road crack recognition model includes: Using the convolutional neural network algorithm to train the model with crack data; Using an independent test set to test and validate the trained model.

5. The recognition and analysis method based on artificial intelligence according to claim 1, characterized in that: The step of evaluating and analyzing the detected cracks includes: Measuring the length, width, depth, and density parameters of the cracks; Setting preset parameter thresholds to determine whether the crack parameters exceed the thresholds; Recording the crack image information for subsequent use.

6. The identification and analysis method based on artificial intelligence according to claim 5, characterized in that: The step of measuring the length, width, depth, and density parameters of the cracks includes: Measuring the length, width, depth, and density parameters of each detected crack. Using computer vision and image processing techniques, through scale calibration and pixel calculation in the image, the actual size of the crack is obtained. For example, based on the known scale information in the image, the pixel length of the crack is calculated, and through the proportional relationship with the actual length, the actual length of the crack is obtained. For the depth of the crack, a non-contact imaging measurement method is adopted: using a high-resolution camera or a laser imaging system to image the road crack, and then using image processing and computer vision techniques to analyze the brightness or texture changes in the image to estimate the depth of the crack. The crack density refers to the number of cracks per unit area. By counting the cracks in the image, the density information of the cracks can be obtained. Using image segmentation and image processing techniques, the crack area is separated from the background, and the cracks are analyzed and counted. The calculation of the crack density can provide the overall distribution of the road surface cracks.

7. An identification and analysis method based on artificial intelligence according to claim 5, characterized in that: The step of recording the crack image information for subsequent use includes: Record and store the crack parameters obtained from measurement and statistics for subsequent data analysis and maintenance plan formulation. At the same time, establish a database or file system to store the parameter information of each detected crack, as well as the corresponding image data and timestamps. This can achieve long-term monitoring and analysis of the road crack condition, providing a decision-making basis for road maintenance and repair. Through the above steps, it is possible to conduct detailed parameter measurement and statistical analysis of the detected cracks, judge the severity of the cracks according to preset thresholds, and achieve automated crack information reporting and repair arrangement, thereby improving the efficiency and accuracy of road maintenance.

8. An identification and analysis method based on artificial intelligence according to claim 1, characterized in that: The steps for crack detection result display and evaluation include: The results of crack detection and analysis are presented to the administrator in a visual form; Conduct an assessment and prediction of the road service life and maintenance requirements.

9. An identification and analysis method based on artificial intelligence according to claim 8, characterized in that: The steps for conducting an assessment and prediction of the road service life and maintenance requirements include: Based on the type and degree of the cracks, the system can conduct an assessment and prediction of the road service life. Based on the existing road materials, design standards, and historical data, through comprehensive analysis of the crack parameters and other relevant factors, the system estimates the remaining service life of the road and provides corresponding suggestions and maintenance requirements. This will help the administrator formulate a reasonable road maintenance plan and budget to extend the road service life and ensure traffic safety.

10. An artificial intelligence-based recognition and analysis system, characterized in that: The system includes: An image acquisition and annotation module for acquiring road crack images and performing crack annotation; A crack detection and recognition module for establishing a detection model to achieve the recognition of road cracks; A crack analysis and evaluation module for analyzing the detected cracks and evaluating the pavement life.

Images