Circuit board drawing restoration method and device based on intelligent scanning

Abstract

The invention belongs to the technical field of circuit board design, and discloses a circuit board drawing restoration method based on intelligent scanning, which comprises the following steps: generating three-dimensional coordinates according to parameters and scanning results of an entity circuit board, identifying element composition of the entity circuit board, and extracting feature data of the entity circuit board; according to the feature data, training through a deep learning model to obtain a trained model; inputting parameters and element composition of an entity circuit board of a to-be-restored drawing into the trained model to generate a corresponding schematic diagram or layout; calling an electric appliance rule of the EDA system to check and adjust the generated schematic diagram or layout until the electric appliance rule is met; the invention further discloses a circuit board drawing restoring device based on intelligent scanning. The method is suitable for restoring the schematic diagram or layout of the circuit board according to the entity circuit board, the design drawing of the entity circuit board is accurately and rapidly restored through image recognition and the deep learning model, and time and energy of design workers are saved.

Description

Technical Field

[0001] The invention belongs to the technical field of circuit board design, and relates to a circuit board drawing restoration method and device based on intelligent scanning. Background Art

[0002] Circuit board design is a very complicated task. First, you need to design the circuit board schematic or layout in the EDA system, then design the circuit board according to the circuit board schematic or layout, and then produce and process the circuit board based on the circuit board design to get the physical circuit board. However, when the circuit board schematic or layout is damaged or lost, engineers need to restore the drawings step by step based on the physical circuit board, which requires a lot of energy and time. Therefore, it is of great significance to design a method to accurately and quickly restore the circuit board drawings by intelligent scanning to identify the physical circuit board. Summary of the invention

[0003] The purpose of the present invention is to provide a circuit board drawing restoration method based on intelligent scanning, which is intended to replace manual restoration of circuit board schematics or layouts, saving the time and energy of designers; Another object of the present invention is to provide a circuit board drawing restoration device based on intelligent scanning.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is as follows: A circuit board drawing restoration method based on intelligent scanning comprises the following steps: S1. Enter the parameters of the physical circuit board in the EDA system; S2. Use a scanning device to scan the physical circuit board and enter the scanning results into the EDA system; S3, generating a three-dimensional coordinate according to the input parameters of the physical circuit board and the scanning result, identifying the component composition of the physical circuit board, and extracting and storing the geometric features, texture features and contextual features data of the physical circuit board according to the component composition; S4. According to the extracted feature data, a deep learning model is trained to obtain a trained deep learning model; S5. Input the parameters and component composition of the physical circuit board to be restored into the trained deep learning model to generate a corresponding schematic or layout; S6. Call the electrical rules of the EDA system to check and adjust the generated schematic or layout until it meets the electrical rules.

[0005] As a limitation, in step S1, the parameters of the physical circuit board include the type of circuit board, the material used, the number of layers, the current capacity, the impedance, the voltage, the size of the circuit board and the thickness of the copper foil.

[0006] As a second limitation, in step S2, the scanning device is a camera.

[0007] As a third limitation, in step S3, the components of the physical circuit board include packages, pads, vias, lines, text, and various graphics; Geometric features include the element's outline, area, perimeter, aspect ratio, eccentricity, boundary features, and area moments; Texture features include local binary patterns and gray-level co-occurrence matrices; The context features include the coordinate positions of components on the physical circuit board, the corresponding relationships among the components, and the network connections among the components.

[0008] As a fourth limitation, in step S4, the deep learning model is a support vector machine SVM, a random forest RF or a convolutional neural network.

[0009] The present invention also provides a circuit board drawing restoration device based on intelligent scanning, comprising: Data preparation and processing module, used to enter the parameters of the physical circuit board into the EDA system; Intelligent scanning module, used to scan the physical circuit board using a scanning device and enter the scanning results into the EDA system; The feature data extraction module is used to generate a three-dimensional coordinate according to the parameters of the physical circuit board and the scanning results, identify the component composition of the physical circuit board, and extract and store the geometric features, texture features and contextual features of the physical circuit board according to the component composition; The model training module is used to train the deep learning model according to the extracted feature data to obtain a trained deep learning model; The drawing generation module is used to input the parameters and component composition of the physical circuit board to be restored into the trained deep learning model to generate the corresponding schematic or layout; The check and adjustment module is used to call the electrical rules of the EDA system to check and adjust the generated schematic or layout until it meets the electrical rules.

[0010] Due to the adoption of the above technical solution, the present invention has the following technical advances compared with the prior art: (1) The present invention identifies the component composition of the circuit board based on the parameters of the physical circuit board and the scanning results, and performs feature extraction. After training through a deep learning model, the efficiency of restoration is improved; (2) The present invention uses image recognition and deep learning models to accurately and quickly restore the schematic diagram or layout of the physical circuit board, saving the time and energy of designers; (3) The present invention recognizes various circuit board types through intelligent scanning and trains through a deep learning model. Depending on the complexity of the circuit board types, it takes only ten minutes. Compared with manual recognition, this method saves more than 95% of the time and has better recognition and restoration effects.

[0011] In summary, the present invention is suitable for restoring a circuit board schematic or layout based on a physical circuit board, and can accurately and quickly restore the design drawings of the physical circuit board through image recognition and deep learning models, saving time and energy of designers. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 Shown is a flow chart of the method of embodiment 1 of the present invention; Figure 2 Shown is a device block diagram of embodiment 2 of the present invention. DETAILED DESCRIPTION

[0013] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation modes in conjunction with the accompanying drawings.

[0014] Embodiment A circuit board drawing restoration method based on intelligent scanning like Figure 1 As shown, this embodiment is a circuit board drawing restoration method based on intelligent scanning, comprising the following steps: S1. Enter the parameters of the physical circuit board in the EDA system; In this step, the parameters of the physical circuit board include the type of circuit board, the material used, the number of layers, the current capacity, the impedance, the voltage, the size of the circuit board and the thickness of the copper foil; S2. Scan the physical circuit board using a scanning device and enter the scanning result into the EDA system; the scanning device may be a camera; S3, generating a three-dimensional coordinate according to the input parameters of the physical circuit board and the scanning result, identifying the component composition of the physical circuit board, and extracting and storing the geometric features, texture features and contextual features data of the physical circuit board according to the component composition; In this step, the components of the physical circuit board include packages, pads, vias, lines, text, and various graphics; wherein the various graphics include circles, squares, triangles, rectangles, and irregular graphics, etc.; Geometric features include the outline, area, perimeter, aspect ratio, eccentricity, boundary features (such as boundary points where grayscale changes suddenly from dark to bright in SAR images) and regional moments (describes the regional shape characteristics of the target based on the relationship between the shape outline of the target and its own center of gravity); Texture features include local binary patterns and gray-level co-occurrence matrices; The context features include the coordinate positions of components on the physical circuit board, the corresponding relationships between components, and the network connections between components; S4. According to the extracted feature data, a deep learning model is trained to obtain a trained deep learning model; In this step, the deep learning model can use support vector machine SVM, random forest RF or convolutional neural network; S5. Input the parameters and component composition of the physical circuit board to be restored into the trained deep learning model to generate a corresponding schematic or layout; S6. Call the electrical rules of the EDA system to check and adjust the generated schematic or layout until it meets the electrical rules.

[0015] Embodiment 2 A circuit board drawing restoration device based on intelligent scanning like Figure 2 As shown, this embodiment is a block diagram of a circuit board drawing restoration device based on intelligent scanning provided according to embodiment 1, and the device includes: Data preparation and processing module, used to enter the parameters of the physical circuit board into the EDA system; Intelligent scanning module, used to scan the physical circuit board using a scanning device and enter the scanning results into the EDA system; The feature data extraction module is used to generate a three-dimensional coordinate according to the parameters of the physical circuit board and the scanning results, identify the component composition of the physical circuit board, and extract and store the geometric features, texture features and contextual features of the physical circuit board according to the component composition; The model training module is used to train the deep learning model according to the extracted feature data to obtain a trained deep learning model; The drawing generation module is used to input the parameters and component composition of the physical circuit board to be restored into the trained deep learning model to generate the corresponding schematic or layout; The check and adjustment module is used to call the electrical rules of the EDA system to check and adjust the generated schematic or layout until it meets the electrical rules.

[0016] Among them, the circuit board drawing restoration device based on intelligent scanning provided in this embodiment only uses the division of the above-mentioned functional modules as an example when performing data processing. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed.

[0017] It should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention is described in detail with reference to the above embodiments, those skilled in the art can still modify the technical solutions described in the above embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A circuit board drawing restoration method based on intelligent scanning, characterized in that: The following steps are involved: S1. Enter the parameters of the physical circuit board in the EDA system; S2. Use a scanning device to scan the physical circuit board and enter the scanning results into the EDA system; S3, generating a three-dimensional coordinate according to the input parameters of the physical circuit board and the scanning result, identifying the component composition of the physical circuit board, and extracting and storing the geometric features, texture features and contextual features data of the physical circuit board according to the component composition; S4. According to the extracted feature data, a deep learning model is trained to obtain a trained deep learning model; S5. Input the parameters and component composition of the physical circuit board to be restored into the trained deep learning model to generate a corresponding schematic or layout; S6. Call the electrical rules of the EDA system to check and adjust the generated schematic or layout until it meets the electrical rules.

2. The circuit board drawing restoration method based on intelligent scanning according to claim 1 is characterized in that: In step S1, the parameters of the physical circuit board include circuit board type, material used, number of layers, current capacity, impedance, voltage, circuit board size and copper foil thickness.

3. The circuit board drawing restoration method based on intelligent scanning according to claim 1 is characterized in that: In step S2, the scanning device is a camera.

4. The circuit board drawing restoration method based on intelligent scanning according to claim 1 is characterized in that: In step S3, the components of the physical circuit board include packages, pads, vias, lines, text, and various graphics; Geometric features include the element's outline, area, perimeter, aspect ratio, eccentricity, boundary features, and area moments; Texture features include local binary patterns and gray-level co-occurrence matrices; The context features include the coordinate positions of components on the physical circuit board, the corresponding relationships among the components, and the network connections among the components.

5. The circuit board drawing restoration method based on intelligent scanning according to claim 1, characterized in that: In step S4, the deep learning model is a support vector machine SVM, a random forest RF or a convolutional neural network.

6. A circuit board drawing restoration device based on intelligent scanning, characterized in that: include: Data preparation and processing module, used to enter the parameters of the physical circuit board into the EDA system; Intelligent scanning module, used to scan the physical circuit board using a scanning device and enter the scanning results into the EDA system; The feature data extraction module is used to generate a three-dimensional coordinate according to the parameters of the physical circuit board and the scanning results, identify the component composition of the physical circuit board, and extract and store the geometric features, texture features and contextual features of the physical circuit board according to the component composition; The model training module is used to train the deep learning model according to the extracted feature data to obtain a trained deep learning model; The drawing generation module is used to input the parameters and component composition of the physical circuit board to be restored into the trained deep learning model to generate the corresponding schematic or layout; The check and adjustment module is used to call the electrical rules of the EDA system to check and adjust the generated schematic or layout until it meets the electrical rules.

Images