CN117387517B - Digital instrument panel quality detection method and system - Google Patents
Digital instrument panel quality detection method and system Download PDFInfo
- Publication number
- CN117387517B CN117387517B CN202311390359.XA CN202311390359A CN117387517B CN 117387517 B CN117387517 B CN 117387517B CN 202311390359 A CN202311390359 A CN 202311390359A CN 117387517 B CN117387517 B CN 117387517B
- Authority
- CN
- China
- Prior art keywords
- instrument panel
- quality
- digital instrument
- appearance
- image information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 139
- 238000009826 distribution Methods 0.000 claims abstract description 72
- 238000012216 screening Methods 0.000 claims abstract description 27
- 238000000034 method Methods 0.000 claims abstract description 14
- 239000013598 vector Substances 0.000 claims description 27
- 238000013527 convolutional neural network Methods 0.000 claims description 13
- 238000003064 k means clustering Methods 0.000 claims description 11
- 238000007689 inspection Methods 0.000 claims description 7
- 238000004140 cleaning Methods 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 5
- 238000010801 machine learning Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000013135 deep learning Methods 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 4
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000002349 favourable effect Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000013077 scoring method Methods 0.000 description 1
- 238000011524 similarity measure Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/24—Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8854—Grading and classifying of flaws
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Chemical & Material Sciences (AREA)
- Biomedical Technology (AREA)
- Immunology (AREA)
- Biochemistry (AREA)
- Databases & Information Systems (AREA)
- Analytical Chemistry (AREA)
- Medical Informatics (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Pathology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Quality & Reliability (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of intelligent detection, in particular to a quality detection method of a digital instrument panel, which comprises the following steps: image acquisition is carried out on the outline appearance of the digital instrument panel, and outline appearance image information is obtained; image acquisition is carried out on the distribution appearance of the digital instrument panel, and distribution appearance image information is obtained; carrying out appearance quality classification on the digital instrument panel based on the outline appearance image information and the distribution appearance image information to obtain a quality classification result; determining a quality screening node according to the quality classification result, carrying out multi-dimensional performance detection on a digital instrument panel passing through the quality screening node, and generating a multi-dimensional performance detection data set; and (3) establishing a quality detection model, and inputting the multidimensional performance detection data set into the quality detection model to obtain a quality detection result. By the method, automatic factory parameter detection is realized, a process of manually detecting a plurality of parameters is replaced, the possibility of parameter detection errors is reduced, and the quality of products and the efficiency of parameter detection are improved.
Description
Technical Field
The invention relates to the technical field of intelligent detection, in particular to a digital instrument panel quality detection method and system.
Background
The digital instrument panel is widely applied, especially becomes an indispensable part in the current electric vehicle manufacturing field, is beneficial to improving the safety and easily acquiring the driving vehicle state information, ensures the driving environment safe for the driver, and can display more abundant information and functions compared with the traditional mechanical instrument panel.
The factory detection is an important step and a necessary step of the assembly and marketing of the digital instrument panel, the existing factory detection usually involves manual detection of various factory parameters, and due to the complicated manual intervention and multiple detections of the same detection target, errors may exist in the final result of parameter detection, so that the product quality is reduced, the detection items are mutually independent, and the detection of the complete equipment needs manual information recording, is qualified or not, and greatly consumes labor cost.
The information disclosed in this background section is only for enhancement of understanding of the general background of the disclosure and is not to be taken as an admission or any form of suggestion that this information forms the prior art that is well known to a person skilled in the art.
Disclosure of Invention
The invention provides a digital instrument panel quality detection method and a digital instrument panel quality detection system, which can effectively solve the problems in the background technology.
In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:
a digital instrument panel quality detection method, the method comprising:
image acquisition is carried out on the outline appearance of the digital instrument panel, and outline appearance image information is obtained;
image acquisition is carried out on the distribution appearance of the digital instrument panel, and distribution appearance image information is obtained;
performing appearance quality classification on the digital instrument panel based on the outline appearance image information and the distribution appearance image information to obtain a quality classification result;
determining a quality screening node according to the quality classification result, carrying out multi-dimensional performance detection on the digital instrument panel passing through the quality screening node, and generating a multi-dimensional performance detection data set;
and establishing a quality detection model, and inputting the multidimensional performance detection data set into the quality detection model to obtain a quality detection result.
Further, image acquisition is performed on the outline appearance of the digital instrument panel to obtain outline appearance image information, including:
establishing a three-dimensional acquisition space according to the outline appearance of the digital instrument panel;
determining the position of the contour boundary of the digital instrument panel in the three-dimensional acquisition space;
and scanning the outline appearance of the digital instrument panel to obtain outline appearance image information.
Further, determining a position of a contour boundary of the digital instrument panel in the three-dimensional acquisition space includes:
establishing a three-dimensional coordinate system in the three-dimensional acquisition space;
setting the density of the three-dimensional coordinate system according to the outline of the digital instrument panel;
setting a fixed point as a first tag point for the outline of the digital instrument panel;
and coinciding the first label point with the origin of the three-dimensional coordinate system, wherein the first label point is arranged to ensure that the outline appearance of the digital instrument panel is integrally located in a first quadrant of the three-dimensional coordinate system.
Further, image acquisition is performed on the distribution appearance of the digital instrument panel to obtain distribution appearance image information, including:
determining a position of a power distribution screen of the digital instrument panel;
selecting an acquisition angle perpendicular to the power distribution screen, and determining an acquisition distance;
and carrying out brightness recognition on the acquired distribution screen image to obtain the distribution appearance image information.
Further, the brightness recognition of the collected power distribution screen image comprises the following steps:
converting the distribution screen image into a gray value image;
setting the gray value of the non-luminous area of the power distribution screen image to 0;
constructing a convolutional neural network, and constructing a convolutional layer of the convolutional neural network based on a brightness preset value of the power distribution screen image;
converting the gray value image into a matrix for the convolutional neural network to recognize;
and carrying out brightness recognition on the distribution screen image through the convolutional neural network.
Further, the digital instrument panel is subjected to appearance quality classification based on the outline appearance image information and the distribution appearance image information, and a quality classification result is obtained, including:
preprocessing the outline appearance image information and the distribution appearance image information respectively;
extracting proper characteristics from the processed outline appearance image information and the processed distribution appearance image information, wherein the extracted characteristics can be used for a K-means clustering algorithm;
converting the result of extracting the features into feature vectors, and ensuring the consistency of the dimensions of the feature vectors;
and clustering the feature vectors by using the K-means clustering algorithm, and obtaining the quality classification result.
Further, clustering the feature vectors using the K-means clustering algorithm includes:
s1: initializing a cluster center, wherein the cluster center is a concentration point of the feature vector;
s2: calculating a distance between the feature vector and the cluster center;
s3: distributing the corresponding cluster centers to the feature vectors according to the calculated distance;
s4: updating the cluster center according to all the allocated feature vectors;
s5: setting a stopping condition, and repeating the steps S2-S4 according to the stopping condition.
Further, performing multi-dimensional performance detection on the digital dashboards passing through the quality screening node, comprising:
the quality screening node judges the quality classification result and screens out the digital instrument panel with the appearance quality passing through the standard;
determining a detection item of the multi-dimensional performance detection, and detecting the performance of the digital instrument panel according to the detection item;
and generating the multi-dimensional performance detection data set by using the multi-dimensional performance detection result of the single digital instrument panel.
Further, establishing a quality detection model, including:
collecting historical detection data information of the digital instrument panel;
performing data cleaning on the historical detection data information;
deep learning is carried out on the history detection data information after cleaning based on a machine learning model;
and establishing the quality detection model, inserting a feature extraction layer, and extracting the data features in the multi-dimensional performance detection data set.
A digital instrument panel quality inspection system, the system comprising:
the contour image acquisition module is used for acquiring images of contour appearances of the digital instrument panel to obtain contour appearance image information;
the distribution image acquisition module is used for acquiring images of distribution appearance of the digital instrument panel to obtain distribution appearance image information;
the appearance quality classification module is used for carrying out appearance quality classification on the digital instrument panel based on the outline appearance image information and the distribution appearance image information to obtain a quality classification result;
the quality screening module is used for determining quality screening nodes according to the quality classification result, carrying out multidimensional performance detection on the digital instrument panels passing through the quality screening nodes, and generating a multidimensional performance detection data set;
and the quality detection module is used for establishing a quality detection model, inputting the multidimensional performance detection data set into the quality detection model and obtaining a quality detection result.
By the technical scheme of the invention, the following technical effects can be realized:
the invention realizes automatic factory parameter detection, comprehensively evaluates the appearance quality detection and the multidimensional performance quality detection of the digital instrument panel, replaces the process of manually detecting a plurality of parameters and evaluating the quality detection, effectively reduces the possibility of parameter detection errors, is beneficial to improving the quality of products and the efficiency of parameter detection, and effectively reduces the product residue rate flowing into the market.
The foregoing description is only an overview of the technical solutions of the present application, and may be implemented according to the content of the specification in order to make the technical means of the present application more clearly understood, and in order to make the above-mentioned and other objects, features and advantages of the present application more clearly understood, the following detailed description of the present application will be given.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments described in the present invention, and other drawings may be obtained according to the drawings without inventive effort to those skilled in the art.
FIG. 1 is a flow chart of a method for detecting quality of a digital instrument panel;
FIG. 2 is a flow chart for obtaining outline appearance image information;
FIG. 3 is a flow chart illustrating determining the contour boundary of a digital instrument panel;
FIG. 4 is a flow chart of obtaining quality classification results;
FIG. 5 is a schematic flow chart of clustering feature vectors using a K-means clustering algorithm;
FIG. 6 is a schematic diagram of a digital instrument panel quality inspection system.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.
Example 1
As shown in fig. 1, the present application provides a digital instrument panel quality detection method, which includes:
s100: image acquisition is carried out on the outline appearance of the digital instrument panel, and outline appearance image information is obtained; the method comprises the steps of firstly selecting image acquisition equipment, carrying out image acquisition on the outline appearance of a digital instrument panel, aiming at carrying out image recognition on the appearance shell part of the instrument panel, and knowing the production quality condition of a shell for later comparison and reference.
S200: image acquisition is carried out on the distribution appearance of the digital instrument panel, and distribution appearance image information is obtained; the purpose of the steps is that the power distribution appearance is displayed on a screen of the instrument panel in the power-on state, and appearance images of the instrument panel are respectively collected through control in the power-on state and the power-off state, so that the appearance production quality of the instrument panel can be more comprehensively reflected.
S300: carrying out appearance quality classification on the digital instrument panel based on the outline appearance image information and the distribution appearance image information to obtain a quality classification result; in the step, the instrument panels of a single unit are not used as detection units, the instrument panels of a certain number of batches are collected to form quality classification, and the quality classification can be carried out on the instrument panels of the same batch in a mode of attaching electronic tags to the single instrument panels, so that the quality classification is carried out on the digital instrument panels of the same batch in a mode of establishing a queue by single attribute, whether the single attribute reaches the most basic standard or not is judged, namely whether defective products exist or not is judged, and the quality can be further classified on the condition that the qualified attribute is met; it should be noted that, in the appearance quality, in this embodiment, two single attributes are adopted to evaluate the appearance quality of the instrument panel, and the two attributes are respectively evaluated in separate queues, which may be determined in quality grade by a comprehensive scoring method, different evaluation weights of different attributes may be given to the comprehensive scoring according to corresponding product features, the weights may be equally divided, and finally, the comprehensive evaluation returns to the same instrument panel through an electronic tag, so as to generate a quality classification result.
S400: determining a quality screening node according to the quality classification result, carrying out multi-dimensional performance detection on a digital instrument panel passing through the quality screening node, and generating a multi-dimensional performance detection data set; the quality screening node here may be a qualified quality punctuation or a quality grade dividing punctuation, and the quality screening node meeting the own requirements is formulated according to different working and subsequent processed digital instrument panels, but in this embodiment, instead of only one quality screening node, a plurality of quality screening nodes may be set to divide the quality layer by layer, and according to the result of dividing the quality, other performances of the digital instrument panels are detected by taking a single digital instrument panel as a unit, and these performances may include vibration detection, high temperature resistance, corrosion resistance detection, and the like, and multiple results of the same instrument panel are generated into a multi-dimensional performance detection data set.
S500: and (3) establishing a quality detection model, and inputting the multidimensional performance detection data set into the quality detection model to obtain a quality detection result. In the step, the collected multi-dimensional performance detection data set is input into a quality detection model, the multi-dimensional performance detection data set corresponds to the label of the digital instrument panel and the appearance quality detection result, the quality detection model makes quality evaluation on all detection data in the multi-dimensional performance detection data set, and then the final quality detection result is obtained on the factory parameters of the digital instrument panel according to the previously divided quality detection result about the appearance.
The invention realizes automatic factory parameter detection, comprehensively evaluates the appearance quality detection and the multidimensional performance quality detection of the digital instrument panel, replaces the process of manually detecting a plurality of parameters and evaluating the quality detection, effectively reduces the possibility of parameter detection errors, is beneficial to improving the quality of products and the efficiency of parameter detection, and effectively reduces the product residue rate flowing into the market.
Further, as shown in fig. 2, image acquisition is performed on the outline appearance of the digital instrument panel to obtain outline appearance image information, including:
s110: establishing a three-dimensional acquisition space according to the outline appearance of the digital instrument panel;
s120: determining the position of the contour boundary of the digital instrument panel in a three-dimensional acquisition space;
s130: and scanning the outline appearance of the digital instrument panel to obtain outline appearance image information.
On the basis of the above embodiment, a three-dimensional acquisition space is established to draw points on the wide appearance of the digital instrument panel, and it is required to pay attention to the establishment of the three-dimensional acquisition space based on the range of the outer contour of the digital instrument panel, that is, considering the distance between the boundaries of the actual shell contour, in order to consider the problems of the difficulty and the calculation amount of the acquisition of the subsequent image points at the beginning of the establishment of the acquisition space, which is helpful to improve the overall acquisition and detection speed, in the process of actually establishing the acquisition space, the outer contour of the digital instrument panel in various patterns can be considered to establish the appearance contour of the digital instrument panel which can compatibly detect a plurality of batches, after the three-dimensional acquisition space is established, a plurality of acquisition instruments are arranged in the three-dimensional acquisition space, in particular, a laser scanner or a structured light scanner can be used to scan the space to generate point cloud data, and then the required coordinate points are extracted from the point cloud data, and the outline appearance image information is obtained.
Further, as shown in fig. 3, determining the position of the contour boundary of the digital instrument panel in the three-dimensional acquisition space includes:
s111: establishing a three-dimensional coordinate system in a three-dimensional acquisition space;
s112: setting the density of a three-dimensional coordinate system according to the outline of the digital instrument panel;
s113: setting a fixed point as a first tag point on the outline of the digital instrument panel;
s114: and coinciding the first label point with the origin of the three-dimensional coordinate system, wherein the first label point is arranged to ensure that the outline appearance of the digital instrument panel is wholly in the first quadrant of the three-dimensional coordinate system.
On the basis of the above embodiment, the density of the coordinate system is set according to the contour features of the specific digital instrument panel, the higher the density is, the finer the contour is expressed, but more data points are needed, the density is usually expressed in terms of points per unit distance, for example, per centimeter, while the density of coordinate points is set, attention is paid to selecting a fixed point on one contour as a first label point, the consistency and the accuracy in subsequent calculation can be ensured by selecting the fixed point as a reference, the first label point can be used as a known point in the coordinate system for positioning and calibration, the first label point is selected to coincide with the origin of the three-dimensional coordinate system, the subsequent calculation and operation can be simplified by aligning the first label point with the origin, and the accuracy of the measurement result can be ensured; in addition, the first label point is required to ensure that the outline appearance of the digital instrument panel is entirely located in the first quadrant of the three-dimensional coordinate system, so that the subsequent calculation processing is convenient and meets the expectations.
Further, image acquisition is carried out on the distribution appearance of the digital instrument panel to obtain distribution appearance image information, and the method comprises the following steps:
determining a position of a power distribution screen of the digital instrument panel;
selecting an acquisition angle perpendicular to a power distribution screen, and determining an acquisition distance;
and carrying out brightness recognition on the acquired distribution screen image to obtain distribution appearance image information.
In this embodiment, the power distribution screen refers to an area with a certain brightness and color for displaying a plurality of information and functions of the electric vehicle in the electrified state of the instrument panel, and in general, the display design of the instrument panel of the electric vehicle has a centralization property, and the difference between the power distribution screen and the power distribution screen is that only plane image information is required to be collected for collecting the power distribution appearance image information, and the embodiment selects to identify the brightness in the power distribution screen image and obtain the quality information of the digital instrument panel in the power distribution state through the feedback of the brightness.
Further, the brightness recognition of the collected power distribution screen image comprises the following steps:
converting the distribution screen image into a gray value image;
setting the gray value of a non-luminous area of the power distribution screen image to 0;
constructing a convolutional neural network, and constructing a convolutional layer of the convolutional neural network based on a brightness preset value of the power distribution screen image;
converting the gray value image into a matrix for convolutional neural network identification;
and carrying out brightness recognition on the distribution screen image through a convolutional neural network.
On the basis of the above embodiment, after the image is converted into the gray image, each pixel point does not contain color information any more, but only contains brightness information, and each pixel value in the gray image represents brightness intensity at a corresponding position, usually between 0 (black) and 255 (white), 0 represents darkest, 255 represents brightest, and setting the gray value of the non-brightness area to 0 facilitates accurate identification of the gray image by the convolution layer of the convolution neural network, so that brightness judgment errors formed by interference of light reflection and high brightness points are avoided; using convolutional neural networks and constructing convolutional layers to identify gray scale images, first, the convolutional layers can extract local features of the image, such as edges and textures, to better understand the structure and content of the image. Second, the parameter sharing mechanism reduces the amount of parameters of the network, reduces the risk of overfitting, and makes the network more efficient and trainable.
Further, as shown in fig. 4, the appearance quality classification is performed on the digital instrument panel based on the outline appearance image information and the distribution appearance image information, and a quality classification result is obtained, including:
s310: preprocessing outline appearance image information and distribution appearance image information respectively;
s320: extracting proper characteristics from the processed outline appearance image information and the processed distribution appearance image information, wherein the extracted characteristics can be used for a K-means clustering algorithm;
s330: converting the result of extracting the features into feature vectors, and ensuring the consistency of the dimensions of the feature vectors;
s340: and clustering the feature vectors by using a K-means clustering algorithm, and obtaining a quality classification result.
As a preference of the above embodiment, preprocessing includes denoising, enhancing, etc. the image, ensuring accuracy of feature extraction, determining classification standards according to actual requirements, and correspondingly extracting features according to the required classification standards, wherein the features may include shape, texture, color, brightness, etc., and then converting the extracted features into feature vectors with uniform dimensions, ensuring that each feature vector has the same dimensions, so that the K-means clustering algorithm can effectively operate, which usually involves connecting different features together or reducing dimensions by using dimension reduction technology such as principal component analysis.
Further, as shown in fig. 5, clustering feature vectors using a K-means clustering algorithm includes:
s1: initializing a cluster center which is a concentration point of the feature vector;
s2: calculating the distance between the feature vector and the cluster center;
s3: distributing the feature vector to the corresponding cluster center according to the calculated distance;
s4: updating the cluster center according to all the allocated feature vectors;
s5: setting a stopping condition, and repeating the steps S2-S4 according to the stopping condition.
On the basis of the above embodiment, K initial cluster centers are first selected, the initial cluster centers may be selected according to an expected value, or may be manually set according to detection experience, for each data point, the distance between the initial cluster centers and K cluster centers is calculated, usually, euclidean distance or other similarity measure may be used, the data point is allocated to the cluster center closest to the initial cluster center, that is, the data point is allocated to the cluster closest to the initial cluster center, for each cluster, the average value (centroid) of all the data points in the cluster is calculated, this average value is taken as a new cluster center point, the position of the cluster center is updated, the iteration is repeated until a stopping condition is met, the common stopping condition includes that the maximum iteration number is reached, the cluster center does not change significantly any more, or a certain convergence threshold is reached, after the K-means algorithm stops iterating, each data point is allocated to one cluster, and the center point of the cluster has been determined.
Further, the multi-dimensional performance detection of the digital instrument panel passing through the quality screening node comprises:
the quality screening node judges the quality classification result and screens out a digital instrument panel with the appearance quality passing through the standard;
determining a detection item of multi-dimensional performance detection, and detecting the performance of the digital instrument panel according to the detection item;
and generating a multi-dimensional performance detection data set from the multi-dimensional performance detection result of the single digital instrument panel.
In this embodiment, the detection items may include various detection items such as waterproof, anti-vibration and light-reflecting properties, durability, etc., which may be customized according to the specification and use of the electric vehicle instrument panel to ensure the quality of the product. Performance and reliability.
Further, establishing a quality detection model, including:
collecting historical detection data information of a digital instrument panel;
carrying out data cleaning on the historical detection data information;
deep learning is carried out on the history detection data information after cleaning based on a machine learning model;
and a quality detection model is built, a feature extraction layer is inserted, and data features in the multi-dimensional performance detection data set are extracted.
Based on the above embodiment, according to the collected historical digital instrument panel detection data, including multidimensional performance parameters and corresponding quality labels (such as pass and fail), wherein the data is cleaned, including processing missing values, outliers and noise data, to ensure the quality and consistency of the data, then a proper deep learning model is selected to learn the data and obtain the standardization and grading of the decision quality detection, and the model is trained and verified by alternative historical data, so that the accuracy of the model for outputting quality detection results is improved, the practical meaning of the inserted feature extraction layer is that the original data is converted into a form which is favorable for understanding and processing of the machine learning model, and the dimension and complexity of the data are reduced, which is favorable for improving the performance, the interpretability and the generalization capability of the model, so that the quality classification task of the multidimensional performance detection data is better handled.
Embodiment two:
as shown in fig. 6, a digital instrument panel quality detection system, the system comprising:
the contour image acquisition module is used for acquiring images of contour appearances of the digital instrument panel to obtain contour appearance image information;
the power distribution image acquisition module is used for acquiring the power distribution appearance of the digital instrument panel to obtain power distribution appearance image information;
the appearance quality classification module is used for carrying out appearance quality classification on the digital instrument panel based on the outline appearance image information and the distribution appearance image information to obtain a quality classification result;
the quality screening module is used for determining quality screening nodes according to quality classification results, carrying out multidimensional performance detection on a digital instrument panel passing through the quality screening nodes, and generating a multidimensional performance detection data set;
and the quality detection module is used for establishing a quality detection model, inputting the multidimensional performance detection data set into the quality detection model and obtaining a quality detection result.
The adjusting system in the invention can effectively realize the quality detection method of the digital instrument panel, and has the technical effects as described in the embodiment, and the description is omitted here.
Although the present application has been described in connection with specific features and embodiments thereof, it will be apparent that various modifications and combinations can be made without departing from the spirit and scope of the application. Accordingly, the specification and figures are merely exemplary illustrations of the application as defined in the appended claims and are to be construed as covering any and all modifications, variations, combinations, or equivalents that are within the scope of the application. It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the present application and the equivalents thereof, the present application is intended to cover such modifications and variations.
Claims (8)
1. A digital instrument panel quality detection method, the method comprising:
image acquisition is carried out on the outline appearance of the digital instrument panel, and outline appearance image information is obtained;
image acquisition is carried out on the distribution appearance of the digital instrument panel, and distribution appearance image information is obtained;
performing appearance quality classification on the digital instrument panel based on the outline appearance image information and the distribution appearance image information to obtain a quality classification result, including:
preprocessing the outline appearance image information and the distribution appearance image information respectively;
extracting proper characteristics from the processed outline appearance image information and the processed distribution appearance image information, wherein the extracted characteristics are used for a K-means clustering algorithm;
converting the result of extracting the features into feature vectors, and ensuring the consistency of the dimensions of the feature vectors;
clustering the feature vectors by using the K-means clustering algorithm, and obtaining the quality classification result;
determining a quality screening node according to the quality classification result, carrying out multi-dimensional performance detection on the digital instrument panel passing through the quality screening node, and generating a multi-dimensional performance detection data set;
establishing a quality detection model, comprising:
collecting historical detection data information of the digital instrument panel;
performing data cleaning on the historical detection data information;
deep learning is carried out on the history detection data information after cleaning based on a machine learning model;
the quality detection model is built, a feature extraction layer is inserted, and data features in the multi-dimensional performance detection data set are extracted;
and inputting the multi-dimensional performance detection data set into the quality detection model to obtain a quality detection result.
2. The digital instrument panel quality detection method according to claim 1, wherein the image capturing of the outline appearance of the digital instrument panel to obtain outline appearance image information comprises:
establishing a three-dimensional acquisition space according to the outline appearance of the digital instrument panel;
determining the position of the contour boundary of the digital instrument panel in the three-dimensional acquisition space;
and scanning the outline appearance of the digital instrument panel to obtain outline appearance image information.
3. The digital instrument panel quality detection method of claim 2, wherein determining the position of the contour boundary of the digital instrument panel in the three-dimensional acquisition space comprises:
establishing a three-dimensional coordinate system in the three-dimensional acquisition space;
setting the density of the three-dimensional coordinate system according to the outline of the digital instrument panel;
setting a fixed point as a first tag point for the outline of the digital instrument panel;
and coinciding the first label point with the origin of the three-dimensional coordinate system, wherein the first label point is arranged to ensure that the outline appearance of the digital instrument panel is integrally located in a first quadrant of the three-dimensional coordinate system.
4. The digital instrument panel quality inspection method of claim 3, wherein image capturing of the distribution appearance of the digital instrument panel to obtain distribution appearance image information comprises:
determining a position of a power distribution screen of the digital instrument panel;
selecting an acquisition angle perpendicular to the power distribution screen, and determining an acquisition distance;
and carrying out brightness recognition on the acquired distribution screen image to obtain the distribution appearance image information.
5. The method for detecting the quality of the digital instrument panel according to claim 4, wherein the step of recognizing the brightness of the collected power distribution screen image comprises the steps of:
converting the distribution screen image into a gray value image;
setting the gray value of the non-luminous area of the power distribution screen image to 0;
constructing a convolutional neural network, and constructing a convolutional layer of the convolutional neural network based on a brightness preset value of the power distribution screen image;
converting the gray value image into a matrix for the convolutional neural network to recognize;
and carrying out brightness recognition on the distribution screen image through the convolutional neural network.
6. The digital instrument panel quality detection method of claim 1, wherein clustering the feature vectors using the K-means clustering algorithm comprises:
s1: initializing a cluster center, wherein the cluster center is a concentration point of the feature vector;
s2: calculating a distance between the feature vector and the cluster center;
s3: distributing the corresponding cluster centers to the feature vectors according to the calculated distance;
s4: updating the cluster center according to all the allocated feature vectors;
s5: setting a stopping condition, and repeating the steps S2-S4 according to the stopping condition.
7. The digital instrument panel quality inspection method of claim 1, wherein performing a multi-dimensional performance inspection of the digital instrument panel passing through the quality screening node comprises:
the quality screening node judges the quality classification result and screens out the digital instrument panel with the appearance quality passing through the standard;
determining a detection item of the multi-dimensional performance detection, and detecting the performance of the digital instrument panel according to the detection item;
and generating the multi-dimensional performance detection data set by using the multi-dimensional performance detection result of the single digital instrument panel.
8. A digital instrument panel quality inspection system employing the digital instrument panel quality inspection method of claim 1, the system comprising:
the contour image acquisition module is used for acquiring images of contour appearances of the digital instrument panel to obtain contour appearance image information;
the distribution image acquisition module is used for acquiring images of distribution appearance of the digital instrument panel to obtain distribution appearance image information;
the appearance quality classification module is used for carrying out appearance quality classification on the digital instrument panel based on the outline appearance image information and the distribution appearance image information to obtain a quality classification result;
the quality screening module is used for determining quality screening nodes according to the quality classification result, carrying out multidimensional performance detection on the digital instrument panels passing through the quality screening nodes, and generating a multidimensional performance detection data set;
and the quality detection module is used for establishing a quality detection model, inputting the multidimensional performance detection data set into the quality detection model and obtaining a quality detection result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311390359.XA CN117387517B (en) | 2023-10-25 | 2023-10-25 | Digital instrument panel quality detection method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311390359.XA CN117387517B (en) | 2023-10-25 | 2023-10-25 | Digital instrument panel quality detection method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117387517A CN117387517A (en) | 2024-01-12 |
| CN117387517B true CN117387517B (en) | 2024-03-15 |
Family
ID=89435537
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311390359.XA Active CN117387517B (en) | 2023-10-25 | 2023-10-25 | Digital instrument panel quality detection method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117387517B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111915572A (en) * | 2020-07-13 | 2020-11-10 | 青岛大学 | Self-adaptive gear pitting quantitative detection system and method based on deep learning |
| CN114529906A (en) * | 2022-02-10 | 2022-05-24 | 国网河南省电力公司电力科学研究院 | Method and system for detecting abnormity of digital instrument of power transmission equipment based on character recognition |
| CN114742829A (en) * | 2022-06-09 | 2022-07-12 | 张家港市民华塑胶有限公司 | Method and system for detecting defects of rayon |
| CN116129417A (en) * | 2023-01-19 | 2023-05-16 | 电子科技大学 | Digital instrument reading detection method based on low-quality image |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20230143374A1 (en) * | 2021-11-05 | 2023-05-11 | Crystal D. Rhodes | Feature extraction, labelling, and object feature map |
-
2023
- 2023-10-25 CN CN202311390359.XA patent/CN117387517B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111915572A (en) * | 2020-07-13 | 2020-11-10 | 青岛大学 | Self-adaptive gear pitting quantitative detection system and method based on deep learning |
| CN114529906A (en) * | 2022-02-10 | 2022-05-24 | 国网河南省电力公司电力科学研究院 | Method and system for detecting abnormity of digital instrument of power transmission equipment based on character recognition |
| CN114742829A (en) * | 2022-06-09 | 2022-07-12 | 张家港市民华塑胶有限公司 | Method and system for detecting defects of rayon |
| CN116129417A (en) * | 2023-01-19 | 2023-05-16 | 电子科技大学 | Digital instrument reading detection method based on low-quality image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117387517A (en) | 2024-01-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109584248B (en) | Infrared target instance segmentation method based on feature fusion and dense connection network | |
| CN115082683B (en) | Injection molding defect detection method based on image processing | |
| JP6573226B2 (en) | DATA GENERATION DEVICE, DATA GENERATION METHOD, AND DATA GENERATION PROGRAM | |
| EP3776462B1 (en) | System and method for image-based target object inspection | |
| CN113449594B (en) | Multilayer network combined remote sensing image ground semantic segmentation and area calculation method | |
| US9183450B2 (en) | Inspection apparatus | |
| CN111652292B (en) | Similar object real-time detection method and system based on NCS and MS | |
| CN113160192A (en) | Visual sense-based snow pressing vehicle appearance defect detection method and device under complex background | |
| CN113052295B (en) | Training method of neural network, object detection method, device and equipment | |
| CN114548253A (en) | Digital twin model construction system based on image recognition and dynamic matching | |
| CN107423771B (en) | Two-time-phase remote sensing image change detection method | |
| CN111563549B (en) | Medical image clustering method based on multitasking evolutionary algorithm | |
| CN113706496B (en) | Aircraft structure crack detection method based on deep learning model | |
| CN109241932B (en) | Thermal infrared human body action identification method based on motion variance map phase characteristics | |
| CN117387517B (en) | Digital instrument panel quality detection method and system | |
| CN110111307B (en) | Immune system feedback simulation system and method for immune teaching | |
| CN115601747A (en) | Method and system for calculating confluency of adherent cells | |
| Sapkota | Segmentation of coloured point cloud data | |
| CN112102334B (en) | Test data generation method based on segmentation variation | |
| CN114548250A (en) | Mobile phone appearance detection method and device based on data analysis | |
| CN115034577A (en) | Electromechanical product neglected loading detection method based on virtual-real edge matching | |
| CN115294430A (en) | Machine vision rubbish identification and positioning technology based on sensor coupling | |
| Yongchun et al. | Research of license plate character features extraction and recognition | |
| CN116756835B (en) | Template combination design method, device, equipment and storage medium | |
| CN111242047A (en) | Image processing method and apparatus, electronic device, and computer-readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |