CN117635922A - Quality identification method based on router network cable interface - Google Patents
Quality identification method based on router network cable interface Download PDFInfo
- Publication number
- CN117635922A CN117635922A CN202311666486.8A CN202311666486A CN117635922A CN 117635922 A CN117635922 A CN 117635922A CN 202311666486 A CN202311666486 A CN 202311666486A CN 117635922 A CN117635922 A CN 117635922A
- Authority
- CN
- China
- Prior art keywords
- line interface
- area
- network line
- network
- region
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000012937 correction Methods 0.000 claims abstract description 30
- 230000002093 peripheral effect Effects 0.000 claims description 35
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 241001270131 Agaricus moelleri Species 0.000 claims description 2
- 238000012545 processing Methods 0.000 abstract description 5
- 238000005286 illumination Methods 0.000 description 6
- 239000003086 colorant Substances 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- 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
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
-
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- 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]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of data processing, in particular to a quality identification method based on a router network cable interface. The method comprises the following steps: firstly, acquiring a side image of a router, then performing data processing on the acquired side image, and calculating initial enhancement indexes of all subareas of all network cable interface areas; obtaining correction enhancement indexes of all pixel points in each network line interface area based on the initial enhancement indexes of all the subareas, the definition indexes of all the network line interface areas and the initial enhancement indexes of all the subareas of the reference network line interface areas; and obtaining an enhanced image of each network line interface area based on the correction enhancement index of each pixel point in each network line interface area, and further carrying out quality identification on the router network line interface. Therefore, the method provided by the invention adopts an image recognition mode, carries out relevant data processing, carries out quality recognition on the router network cable interface, and improves the quality recognition precision of the router network cable interface.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a quality identification method based on a router network cable interface.
Background
The network interface of the router refers to an interface between the network card and the network, if the network interface is damaged, the network can be interrupted, and the use effect of the router is affected, so that at least one qualified router needs to ensure that the assembled network interface is not damaged. However, in the production process of the network cable interface, the tab feet with the conductive effect are assembled on the network cable interface through a machine, so that certain damage is inevitably caused to part of the network cable interface, but a certain depth exists in the network cable interface, so that the internal brightness of the network cable interface is lower, and further, the internal damage of the network cable interface is difficult to directly identify from the appearance.
When the quality of a network cable interface is identified by the existing image processing-based method, the acquired image of the network cable interface to be detected is simply enhanced, the quality of the network cable interface to be detected is judged based on the gray level difference between the enhanced image of the network cable interface to be detected and the standard image of the network interface, but the gray level of a part of the acquired image of the network cable interface to be detected is lower due to the influence of illumination when the image is acquired, and the part of the acquired image of the network cable interface to be detected is still quite different from the standard image of the network interface after the part of the image is simply enhanced, so that the part of the image is easily misjudged as a defect area, and the quality identification precision of the network cable interface of a router is lower.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide a quality identification method based on a router network cable interface, which adopts the following technical scheme:
the invention provides a quality identification method based on a router network cable interface, which comprises the following steps:
acquiring a side image of a router to be detected, wherein the side image at least comprises two network cable interface areas;
for any of the network interface areas: dividing the network line interface region based on the gray value of the pixel point in the network line interface region to obtain a subarea of the network line interface region; obtaining a peripheral gray value corresponding to the network cable interface region; calculating an initial enhancement index of each sub-region of the network line interface region based on the average gray value of the pixel points in each sub-region of the network line interface region, the peripheral gray value and the area of each sub-region of the network line interface region; obtaining a definition index of the network line interface region based on the initial enhancement index and the distance between the center point of each sub-region of the network line interface region and the center point of the network line interface region;
taking a network line interface region with the highest definition index in the side image as a reference network line interface region, and obtaining a correction enhancement index of each pixel point in each network line interface region based on an initial enhancement index of each sub-region of each network line interface region, a definition index of each network line interface region and an initial enhancement index of each sub-region of the reference network line interface region;
obtaining importance degrees of gray scales in each network line interface region based on correction enhancement indexes of each pixel point in each network line interface region, and constructing an importance degree histogram corresponding to each network line interface region based on the importance degrees; performing histogram equalization on the importance histogram to obtain an enhanced image of each network cable interface region; and inputting the enhanced image into a trained neural network to obtain damaged areas in the network cable interface areas, and carrying out quality identification on the router network cable interfaces to be detected based on the information of the damaged areas.
Preferably, the dividing the network line interface area based on the gray value of the pixel point in the network line interface area to obtain a sub-area of the network line interface area includes:
calculating the variance of the gray values of all the pixel points in the network line interface area according to the gray values of all the pixel points in the network line interface area;
judging whether the variance of the gray values of all pixel points of the network line interface area is smaller than or equal to a gray variance threshold value, if yes, not dividing the network line interface area, and taking the network line interface area as a sub-area; if not, respectively halving the length and the width of the network cable interface area to obtain four small areas, respectively judging whether the variance of the gray values of all the pixel points in each small area is smaller than or equal to a gray variance threshold value, and if so, stopping dividing; if not, halving the length and the width of the corresponding small area, and the like until the variance of the gray value of the pixel point in each small area obtained by dividing is smaller than or equal to the gray variance threshold value, and marking each small area obtained finally as a sub-area.
Preferably, the method for obtaining the peripheral gray value corresponding to the network cable interface area comprises the following steps:
acquiring a sub-region containing the minimum abscissa or the minimum ordinate or the maximum abscissa or the maximum ordinate of the pixel points in the network line interface region, and recording the sub-region as a peripheral sub-region of the network line interface region;
calculating the average gray value of the pixel points in each peripheral subarea according to the gray value of each pixel point in each peripheral subarea; and acquiring the mode of the average gray value corresponding to all peripheral subareas of the network line interface area, and marking the mode as the peripheral gray value corresponding to the network line interface area.
Preferably, the calculating the initial enhancement index of each sub-area of the network line interface area based on the average gray value of the pixel point in each sub-area of the network line interface area, the peripheral gray value and the area of each sub-area of the network line interface area includes:
for any sub-region of the network line interface region:
calculating the average gray value of the pixel points in the subarea according to the gray value of each pixel point in the subarea; calculating the ratio of the average gray value of the pixel points in the subarea to the peripheral gray value corresponding to the network line interface area where the subarea is positioned, and marking the ratio as a first ratio; the difference value between the natural constant 1 and the first ratio is recorded as a first index;
calculating the ratio of the area of the subarea to the total area of the network cable interface area where the subarea is positioned, and recording the ratio as a second ratio;
taking the product of the first index and the second ratio as an initial enhancement index of the subarea.
Preferably, the definition index of the network cable interface area is calculated by adopting the following formula:
wherein Z is i Is the definition index of the ith network cable interface area, d ij N is the distance between the center point of the jth sub-area of the ith network line interface area and the center point of the network line interface area i For the number of subregions of the ith network line interface region, exp () is an exponential function based on a natural constant e, p ij Is the initial enhancement index of the jth sub-region of the ith network interface region.
Preferably, the obtaining the correction enhancement index of each pixel point in each network line interface area based on the initial enhancement index of each sub-area of each network line interface area, the definition index of each network line interface area, and the initial enhancement index of each sub-area of the reference network line interface area includes:
for any pixel point in any network interface area:
acquiring a pixel point at the same position as the pixel point in the reference network interface area, and marking the pixel point as a matching point of the pixel point; calculating the absolute value of the difference value between the initial enhancement index of the subarea where the pixel point is located and the initial enhancement index of the subarea where the matching point of the pixel point is located, and recording the absolute value as a first absolute value; the sum of the natural constant 1 and the first absolute value is recorded as a second index;
calculating the ratio of the definition of the network line interface area where the pixel point is located to the definition of the network line interface area where the matching point of the pixel point is located, and marking the ratio as a third ratio; the difference value between the natural constant 1 and the third ratio is recorded as a third index;
and calculating the product of the initial enhancement index, the second index and the third index of the subarea where the pixel point is located, and taking the product as a correction enhancement index of the pixel point.
Preferably, the obtaining the importance level of each gray level in each network line interface area based on the correction enhancement index of each pixel point in each network line interface area includes:
for any of the network interface areas:
and counting the gray levels in the network line interface area and the correction enhancement indexes of all the pixel points corresponding to the gray levels, and calculating the accumulated sum of the correction enhancement indexes of all the pixel points corresponding to the gray levels in the network line interface area as the importance degree of the corresponding gray levels in the network line interface area.
The invention has at least the following beneficial effects:
1. dividing each network line interface region based on gray values of pixel points in each network line interface region, obtaining the same or similar gray of the pixel points in each sub region, then analyzing each sub region independently, and calculating initial enhancement indexes of each sub region based on average gray values of the pixel points in each sub region, corresponding peripheral gray values and the areas of the sub regions; in consideration of the influence of illumination when images of different network interfaces are acquired, the brightness of the acquired images of the different network interfaces is different, so that the subareas obtained by dividing the different network interfaces are not completely the same, if the subareas are enhanced based on the initial enhancement indexes of the subareas only, some pixel points containing texture information are usually ignored, the pixel points can reflect the quality of the network interface areas, in order to ensure the accuracy of subsequent quality identification, the pixel points are required to be enhanced, the network interface area with the acquired definition index is used as a reference network interface area, and because the reference network interface area is the network interface area with the highest definition in the side image of a router to be detected, the invention corrects the initial enhancement indexes of the pixel points in the network interface areas based on the initial enhancement indexes of the pixel points in the network interface areas, further carries out the correction enhancement indexes of the pixel points in the network interface areas based on the correction enhancement indexes of the pixel points in the network interface areas, and further carries out the correction of the network interface areas on the correction enhancement indexes of the pixel points in the network interface areas, thereby ensuring the accuracy of the network interface area to be detected, and the accuracy of the network interface area is improved, and the accuracy of the network interface quality of the network interface area is guaranteed.
2. The invention acquires the side image of the router to be detected, and simultaneously enhances the images of the plurality of network cable interface areas in the side image of the router to be detected, thereby improving the identification efficiency of the quality of the network cable interface and simultaneously guaranteeing the image quality of the network cable interface, thereby guaranteeing the identification precision of the quality of the network cable interface of the router to be detected.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a quality identification method based on a router network cable interface according to the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the present invention to achieve the preset purpose, the following detailed description is given to a quality recognition method based on a router network cable interface according to the present invention with reference to the accompanying drawings and the preferred 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 following specifically describes a specific scheme of the quality identification method based on the router network cable interface provided by the invention with reference to the accompanying drawings.
A quality identification method embodiment based on a router network cable interface:
the embodiment provides a quality identification method based on a router network cable interface, as shown in fig. 1, and the quality identification method based on the router network cable interface comprises the following steps:
step S1, acquiring a side image of a router to be detected, wherein the side image at least comprises two network cable interface areas.
The quality of the assembled router is identified, after the router is assembled, the routers with multiple rows and multiple columns are orderly arranged together, a camera is opposite to a network interface area on the router for image acquisition, a flash lamp is used for supplementing light, the acquired image is processed, and the quality of a network line interface of the router is identified according to the characteristic information of pixel points in the image.
Firstly, acquiring a side image of a router to be detected by using a camera, wherein the side image comprises a plurality of network cable interface areas to be detected, so that the side image of the router to be detected comprising a plurality of network cable interfaces is obtained; firstly, manually marking each network interface area on the surface of the batch of routers, and obtaining each network interface area on the router only by acquiring the corresponding area in the side image of the router to be detected.
Thus, each network cable interface area in the side image of the router to be detected is obtained.
Step S2, for any network interface area: dividing the network line interface region based on the gray value of the pixel point in the network line interface region to obtain a subarea of the network line interface region; obtaining a peripheral gray value corresponding to the network cable interface region; calculating an initial enhancement index of each sub-region of the network line interface region based on the average gray value of the pixel points in each sub-region of the network line interface region, the peripheral gray value and the area of each sub-region of the network line interface region; and obtaining the definition index of the network line interface region based on the initial enhancement index and the distance between the central point of each sub-region of the network line interface region and the central point of the network line interface region.
In this embodiment, each network cable interface area is first analyzed separately, and each network cable interface area is divided to obtain a plurality of sub-areas, and initial enhancement indexes of the sub-areas are determined according to the areas of the sub-areas in the single network cable interface area and the gray values of the internal pixels.
For the i-th network line interface region:
calculating the average value of the abscissa of each pixel point in the network line interface area, and taking the average value as the abscissa of the central point of the network line interface area; calculating the average value of the ordinate of each pixel point in the network cable interface area, and taking the average value as the ordinate of the central point of the network cable interface area; calculating the variance of the gray values of all the pixel points in the network line interface area according to the gray values of all the pixel points in the network line interface area, setting a gray variance threshold delta, judging whether the variance of the gray values of all the pixel points in the network line interface area is less than or equal to delta, if yes, not dividing the network line interface area, and taking the network line interface area as a sub-area; if not, respectively halving the length and the width of the network cable interface area, namely dividing the network cable interface area into four small areas, then respectively judging whether the variance of gray values of all pixel points in each small area is less than or equal to delta, if so, stopping dividing; if not, analogy to the method, respectively halving the length and the width of the corresponding small area to obtain four small areas corresponding to the corresponding small area; repeating the method until the variance of the gray value of the pixel point in each divided small area is less than or equal to delta, and marking each finally obtained small area as a sub-area. So far, the sub-region corresponding to the network line interface region is obtained. In this embodiment, δ has a value of 3, and in a specific application, the practitioner can set the value of δ by himself.
Because the most edge position in the network cable interface is the shell part of the network cable interface, the part is a flat area, and the colors of the network cable interface are more various, and the situation that the gray value is lower due to the colors of the network cable interface cannot be eliminated, the embodiment uses the colors of the shell areas of the network cable interfaces as the contrast gray value to judge whether each subarea needs gray enhancement. Specifically, acquiring a sub-area of the outermost layer in the network line interface area, namely, a sub-area containing the minimum abscissa or the minimum ordinate or the maximum abscissa or the maximum ordinate of the pixel points in the network line interface area, and recording the sub-area as a peripheral sub-area of the network line interface area; according to the gray values of the pixel points in each peripheral subarea, calculating the average gray value of the pixel points in each peripheral subarea, wherein one peripheral subarea corresponds to one average gray value; and acquiring the mode of the average gray value corresponding to all the peripheral subregions, and marking the mode as the peripheral gray value corresponding to the network line interface region.
The larger the area of a sub-region, the less texture in the sub-region, at which time there may be two situations in the sub-region: one is that no texture exists in the subarea per se, and the other is that the subarea is affected by illumination, so that the brightness is low; therefore, further judgment is needed by combining the gray values of the pixel points in the subareas, namely, the larger the area of the subareas is and the smaller the gray values of the pixel points in the subareas are, the larger the probability that the subareas are the bottom areas with dark self in the network line interfaces is, and the larger the probability that the subareas are reinforced is; the magnitude of the gray value of the pixel point in the subarea can be represented by the ratio of the average gray value of the pixel point in the subarea to the peripheral gray value, and the smaller the ratio is, the larger the difference between the gray value of the pixel point in the subarea and the gray value of the pixel point in the shell area is, namely the smaller the gray value of the pixel point in the subarea is; the size of the area of the sub-area can be characterized by the ratio of the area of the sub-area to the total area of the network line interface area where the sub-area is located, the larger the ratio is, which means that the area of the sub-area is closer to the area of the whole network line interface area, i.e. the area of the sub-area is larger.
For the j-th sub-region of the i-th network line interface region:
calculating the average gray value of the pixel points in the subarea according to the gray value of each pixel point in the subarea; calculating an initial enhancement index of the subarea according to the ratio of the average gray value of the pixel points in the subarea to the peripheral gray value corresponding to the network line interface area where the subarea is located and the ratio of the area of the subarea to the total area of the network line interface area where the subarea is located, namely:
wherein p is ij The initial enhancement index of the jth sub-region of the ith network cable interface region is S, which is the total area of the ith network cable interface region ij An area h of a jth sub-area of the ith network line interface area ij An average gray value H of pixel points in the jth sub-area of the ith network line interface area i And the peripheral gray value corresponding to the ith network cable interface area.
When the ratio of the average gray value of the pixel points in the jth sub-area of the ith network line interface area to the peripheral gray value corresponding to the ith network line interface area is smaller, the larger the gray difference between the jth sub-area and the shell area is, the larger the probability that the jth sub-area is the bottom area of the network line interface is, and the sub-area is required to be enhanced; when the ratio of the area of the jth sub-area of the ith network line interface area to the total area of the ith network line interface area is larger, the larger the duty ratio of the jth sub-area in the whole network line interface area is, the more the sub-area needs to be enhanced; therefore, if the ratio of the average gray value of the pixel point in the jth sub-area of the ith network line interface area to the peripheral gray value corresponding to the ith network line interface area is smaller, and the ratio of the area of the jth sub-area of the ith network line interface area to the total area of the ith network line interface area is larger, the jth sub-area of the ith network line interface area needs to be enhanced, that is, the initial enhancement index of the sub-area is larger.
By adopting the method, the initial enhancement indexes of all the subareas of the ith network cable interface area can be obtained.
The more the line interfaces close to the camera reflect more light in the camera due to the difference of the light quantity directions reflected into the camera in different line interfaces, the clearer the interior of the line interfaces close to the camera is, at this time, if textures at a certain position in the clearer line interfaces are richer, the more detailed information at the position is required to be detected, and for the line interfaces with lower brightness, the more pixel points at the position are required to be enhanced, therefore, the embodiment needs to find the line interface with the clearest interior, and because the more detailed information in the line interfaces is, the smaller the subareas are divided, the more the interior light is, the larger the gray value is, that is, the more the line interfaces with the clearest interior meet the characteristics of smaller area and larger gray value of the subareas, the initial enhancement index corresponding to each subarea is smaller, and because the line interfaces have a certain depth, the initial enhancement index of the subareas close to the middle position of the line interfaces is smaller, and the clear interior of the line interfaces is illustrated. Based on this, in this embodiment, the distance between the center point of each sub-area of the network line interface area and the center point of the network line interface area is taken as the reference weight of the corresponding sub-area, and the definition index of the network line interface area is calculated according to the initial enhancement index of each sub-area of the network line interface area and the distance between the center point of each sub-area of the network line interface area and the center point of the network line interface area, that is:
wherein Z is i Is the definition index of the ith network cable interface area, d ij N is the distance between the center point of the jth sub-area of the ith network line interface area and the center point of the network line interface area i Exp () is an exponential function based on a natural constant e, which is the number of subregions of the i-th network line interface region.
When the distance from the center point of each subarea of the network line interface area to the center point of the network line interface area is closer and the initial enhancement index of each subarea of the network line interface area is smaller, the network line interface area is clearer, namely the definition index of the network line interface area is larger; when the distance from the center point of each subarea of the network line interface area to the center point of the network line interface area is longer, the initial enhancement index of each subarea of the network line interface area is larger, the network line interface area is not clear, namely the definition index of the network line interface area is smaller.
By adopting the method, the definition index of each network cable interface area to be detected in the side image of the router to be detected can be obtained.
And S3, taking the network line interface area with the highest definition index in the side image as a reference network line interface area, and obtaining the correction enhancement index of each pixel point in each network line interface area based on the initial enhancement index of each sub-area of each network line interface area, the definition index of each network line interface area and the initial enhancement index of each sub-area of the reference network line interface area.
In order to ensure the accuracy of the internal quality identification of the network cable interfaces of the router to be detected, the clearer and better the internal parts of the network cable interfaces are, but the initial enhancement indexes only aim at single subareas in the single network cable interface area, and the illumination amounts in different network cable interfaces are different, so that the enhancement indexes of the subareas in the network cable interfaces are further adjusted according to the difference condition between the network cable interfaces of the router to be detected and the sharpest network cable interface.
The size and shape of the network cable interfaces of the same batch are the same, and in step S2, the definition index of each network cable interface area to be detected is obtained, and in this embodiment, the network cable interface area to be detected with the highest definition index is selected as the reference network cable interface area.
For the ith pixel point of the jth sub-area of the ith network line interface area:
acquiring a pixel point at the same position as the pixel point in the reference network interface area, and marking the pixel point as a matching point of the pixel point; calculating the absolute value of the difference value between the initial enhancement index of the sub-region where the pixel point is located and the initial enhancement index of the sub-region where the matching point of the pixel point is located, wherein the larger the absolute value is, the more the pixel point should be enhanced; the initial enhancement index of the subarea where the pixel point is located represents the degree of enhancement required by the subarea where the pixel point is located, and the larger the initial enhancement index of the subarea where the pixel point is located is, the larger the enhancement degree of the subarea where the pixel point is located is; the smaller the ratio of the definition of the network line interface region where the pixel point is located to the definition of the network line interface region where the matching point of the pixel point is located, the greater the difference between the definition of the network line interface region where the pixel point is located and the definition of the network line interface region where the matching point of the pixel point is located, that is, the less the network line interface region where the pixel point is located is clear, the greater the interference of illumination on the network line interface region where the pixel point is located when image acquisition is performed, so that the network line interface region should be enhanced. Based on this, in this embodiment, according to the ratio of the sharpness of the network line interface area where the pixel point is located to the sharpness of the network line interface area where the matching point of the pixel point is located, the initial enhancement index of the sub-area where the pixel point is located, and the absolute value of the difference between the initial enhancement index of the sub-area where the pixel point is located and the initial enhancement index of the sub-area where the matching point of the pixel point is located, the correction enhancement index of the pixel point is calculated, that is:
wherein P is iju Correction enhancement index of the ith pixel point of the jth sub-area of the ith network line interface area, y iju Absolute value of difference value of initial enhancement index of jth sub-region and initial enhancement index of sub-region where matching point of jth pixel point is located, Z i Z1 is the definition index of the ith network cable interface area u The definition of the network line interface area where the matching point of the ith pixel point is located.
Is the ratio of the definition of the network interface area where the pixel point is located to the definition of the network interface area where the matching point of the pixel point is located, +.>Characterizing the overall adjustment quantity of the ith network cable interface area; if->The closer the ratio of the (1) to the (1) is, the closer the definition of the (i) th network line interface area is to the definition of the reference network line interface area, namely, the lower the adjustment degree of the initial enhancement indexes of all the subareas of the (i) th network line interface area is; when the ratio of the definition of the network line interface area where the pixel point is located to the definition of the network line interface area where the matching point of the pixel point is located is smaller, the initial enhancement index of the sub-area where the pixel point is located is larger, and the absolute value of the difference value between the initial enhancement index of the sub-area where the pixel point is located and the initial enhancement index of the sub-area where the matching point of the pixel point is located is larger, the corrected initial enhancement index of the pixel point is larger, namely the corrected enhancement index of the pixel point is larger; when the ratio of the definition of the network line interface area where the pixel point is located to the definition of the network line interface area where the matching point of the pixel point is located is larger, the initial enhancement index of the sub-area where the pixel point is located is smaller, and the absolute value of the difference value between the initial enhancement index of the sub-area where the pixel point is located and the initial enhancement index of the sub-area where the matching point of the pixel point is located is smaller, the corrected initial enhancement index of the pixel point is indicated to be smaller, namely the corrected enhancement index of the pixel point is smaller.
By adopting the method provided by the invention, the correction enhancement index of each pixel point of each network cable interface area of the router to be detected can be obtained. Because the size and the shape of the network cable interfaces of the same batch are the same, when the quality of the network cable interfaces to be detected of the same batch is detected, the brightness of the initially acquired images of different network cable interfaces is different, the pixel points at the same position on different network cable interfaces may be in different subareas, when the matched pixel point of one pixel point is in a smaller subarea, the difference between the pixel point and the initial enhancement index corresponding to the matched pixel point is larger, which means that the pixel point is more likely to contain texture information, in order to ensure the accuracy of recognition, the pixel point is more likely to contain texture information, and in order to ensure the accuracy of recognition, the embodiment adjusts the initial enhancement index of each subarea in the whole network cable interface area, and then adjusts each pixel point in each network cable interface area so as to improve the recognition accuracy of the quality of the subsequent network cable interfaces.
Step S4, obtaining the importance degree of each gray level in each network line interface area based on the correction enhancement index of each pixel point in each network line interface area, and constructing an importance degree histogram corresponding to each network line interface area based on the importance degree; performing histogram equalization on the importance histogram to obtain an enhanced image of each network cable interface region; and inputting the enhanced image into a trained neural network to obtain damaged areas in the network cable interface areas, and carrying out quality identification on the router network cable interfaces to be detected based on the information of the damaged areas.
In step S3, a correction and enhancement index of each pixel point of each network cable interface area is obtained, and then each network cable interface area is enhanced based on the correction and enhancement index of each pixel point of each network cable interface area, so that the quality of the network cable interface to be detected is accurately identified.
For the i-th network line interface region:
the larger the correction enhancement index of each pixel point is, the larger the probability that the corresponding gray level belongs to the internal texture information of the network line interface is, namely the more attention is needed when the corresponding position is the network line interface for quality identification, that is, the more important the gray level of the pixel point is, in order to make the image clearer, the difference between the gray levels at the important positions is needed to be increased, so that the image contrast is improved, and the image quality is enhanced. Therefore, in this embodiment, the accumulated sum of the correction enhancement indexes of each gray level and all pixel points corresponding to each gray level in the network interface area is counted, and the importance level of each gray level in the network interface area is normalized as the importance level of the corresponding gray level, that is, the importance level of each gray level is divided by the accumulated sum of the importance levels of all gray levels in the network interface area. And then, constructing an importance level histogram based on each gray level and the normalized importance level thereof in the network line interface region to obtain an importance level histogram corresponding to the network line interface region, and carrying out histogram equalization on the importance level histogram corresponding to the network line interface region, wherein the i-th network line interface region after equalization is the result after image enhancement on the i-th network line interface region. The method for constructing the histogram and the histogram equalization are both the prior art, and are not described in detail here.
By adopting the method, the enhanced image of each network cable interface area can be obtained.
Because of the diversity of the defects in the network cable interfaces, the quality of each network cable interface is identified by using the quality identification network, the structure of the quality identification network is an Encoder-Decoder structure, the loss function of the network is a cross entropy loss function, the training set of the network is sample images of various network cable interfaces, damaged areas exist in the sample images, the task of the network in the embodiment is classification, and the pixel points in the images need to be divided into two types, so that the corresponding label marking process of the training set is as follows: and (3) inputting the sample image into the network to obtain a normal region and a damaged region by using the single-channel semantic tag, wherein the label of the pixel at the corresponding position belongs to the damaged region as 1, and the label of the pixel belongs to the normal region as 0. The training process of the network is the prior art and will not be described in detail here.
And inputting the enhanced image of each network line area of the router to be detected into a trained quality recognition network to obtain a damaged area in each network line interface area. The larger the area of the damaged area in a certain network cable interface area, the poorer the quality of the network cable interface. And then carrying out quality identification on the router network cable interface to be detected according to the information of the damaged area, wherein the information of the damaged area comprises the number, the area, the position and the like. The practitioner can set an area threshold according to specific conditions, and when the area of the damaged area in the network cable interface area is larger than the area threshold, the quality of the network cable interface area is judged to be inconsistent with the requirement; and when the area of the damaged area in the network cable interface area is smaller than or equal to an area threshold value, judging that the quality of the network cable interface area meets the requirements. If the router to be detected has an unsatisfactory network cable area, judging that the router to be detected has a quality problem. Thus, the quality identification of the router network cable interface to be detected is completed.
Dividing each network line interface region based on gray values of pixel points in each network line interface region, obtaining the same or similar gray of the pixel points in each sub region, then analyzing each sub region independently, and calculating initial enhancement indexes of each sub region based on average gray values of the pixel points in each sub region, corresponding peripheral gray values and areas of each sub region; in consideration of the influence of illumination when images of different network interfaces are acquired, the brightness of the acquired images of the different network interfaces is different, so that the subareas obtained by dividing the different network interfaces are not completely the same, if the subareas are enhanced based on the initial enhancement indexes of the subareas only, some pixel points containing texture information are often ignored, the quality of the network interface areas can be reflected more by the pixel points, in order to ensure the accuracy of subsequent quality identification, the pixel points are required to be enhanced more, the network interface area with the definition index is acquired as a reference network interface area, and because the reference network interface area is the network interface area with the highest definition in the side images of the router to be detected, the embodiment corrects the initial enhancement indexes of the pixel points in the network interface areas based on the initial enhancement indexes of the pixel points in the reference network interface area, further carries out the correction enhancement indexes of the pixel points in the network interface area based on the correction enhancement indexes of the pixel points in the network interface area, and the accuracy of the network interface area is guaranteed to be improved, and the accuracy of the network interface area to be detected is further improved, and the accuracy of the network interface area is guaranteed. According to the method and the device for detecting the network cable interface, the side image of the router to be detected is obtained, the images of the network cable interface areas in the side image of the router to be detected are subjected to image enhancement at the same time, the identification efficiency of the quality of the network cable interface is improved, the image quality of the network cable interface is guaranteed, and therefore the identification precision of the quality of the network cable interface of the router to be detected is guaranteed.
It should be noted that: the foregoing description of the preferred embodiments of the present invention is not intended to be limiting, but rather, any modifications, equivalents, improvements, etc. that fall within the principles of the present invention are intended to be included within the scope of the present invention.
Claims (7)
1. The quality identification method based on the router network cable interface is characterized by comprising the following steps:
acquiring a side image of a router to be detected, wherein the side image at least comprises two network cable interface areas;
for any of the network interface areas: dividing the network line interface region based on the gray value of the pixel point in the network line interface region to obtain a subarea of the network line interface region; obtaining a peripheral gray value corresponding to the network cable interface region; calculating an initial enhancement index of each sub-region of the network line interface region based on the average gray value of the pixel points in each sub-region of the network line interface region, the peripheral gray value and the area of each sub-region of the network line interface region; obtaining a definition index of the network line interface region based on the initial enhancement index and the distance between the center point of each sub-region of the network line interface region and the center point of the network line interface region;
taking a network line interface region with the highest definition index in the side image as a reference network line interface region, and obtaining a correction enhancement index of each pixel point in each network line interface region based on an initial enhancement index of each sub-region of each network line interface region, a definition index of each network line interface region and an initial enhancement index of each sub-region of the reference network line interface region;
obtaining importance degrees of gray scales in each network line interface region based on correction enhancement indexes of each pixel point in each network line interface region, and constructing an importance degree histogram corresponding to each network line interface region based on the importance degrees; performing histogram equalization on the importance histogram to obtain an enhanced image of each network cable interface region; and inputting the enhanced image into a trained neural network to obtain damaged areas in the network cable interface areas, and carrying out quality identification on the router network cable interfaces to be detected based on the information of the damaged areas.
2. The quality identification method based on the router network line interface according to claim 1, wherein the dividing the network line interface area based on the gray value of the pixel point in the network line interface area to obtain the sub-area of the network line interface area includes:
calculating the variance of the gray values of all the pixel points in the network line interface area according to the gray values of all the pixel points in the network line interface area;
judging whether the variance of the gray values of all pixel points of the network line interface area is smaller than or equal to a gray variance threshold value, if yes, not dividing the network line interface area, and taking the network line interface area as a sub-area; if not, respectively halving the length and the width of the network cable interface area to obtain four small areas, respectively judging whether the variance of the gray values of all the pixel points in each small area is smaller than or equal to a gray variance threshold value, and if so, stopping dividing; if not, halving the length and the width of the corresponding small area, and the like until the variance of the gray value of the pixel point in each small area obtained by dividing is smaller than or equal to the gray variance threshold value, and marking each small area obtained finally as a sub-area.
3. The quality identification method based on the router network cable interface according to claim 1, wherein the method for obtaining the peripheral gray value corresponding to the network cable interface region is as follows:
acquiring a sub-region containing the minimum abscissa or the minimum ordinate or the maximum abscissa or the maximum ordinate of the pixel points in the network line interface region, and recording the sub-region as a peripheral sub-region of the network line interface region;
calculating the average gray value of the pixel points in each peripheral subarea according to the gray value of each pixel point in each peripheral subarea; and acquiring the mode of the average gray value corresponding to all peripheral subareas of the network line interface area, and marking the mode as the peripheral gray value corresponding to the network line interface area.
4. The quality recognition method based on a router network line interface according to claim 1, wherein the calculating an initial enhancement index of each sub-area of the network line interface area based on the average gray value of the pixel points in each sub-area of the network line interface area, the peripheral gray value, and the area of each sub-area of the network line interface area comprises:
for any sub-region of the network line interface region:
calculating the average gray value of the pixel points in the subarea according to the gray value of each pixel point in the subarea; calculating the ratio of the average gray value of the pixel points in the subarea to the peripheral gray value corresponding to the network line interface area where the subarea is positioned, and marking the ratio as a first ratio; the difference value between the natural constant 1 and the first ratio is recorded as a first index;
calculating the ratio of the area of the subarea to the total area of the network cable interface area where the subarea is positioned, and recording the ratio as a second ratio;
taking the product of the first index and the second ratio as an initial enhancement index of the subarea.
5. The quality identification method based on the router network cable interface according to claim 1, wherein the definition index of the network cable interface area is calculated by adopting the following formula:
wherein Z is i As a definition index of the i-th network line interface area,d ij n is the distance between the center point of the jth sub-area of the ith network line interface area and the center point of the network line interface area i For the number of subregions of the ith network line interface region, exp () is an exponential function based on a natural constant e, p ij Is the initial enhancement index of the jth sub-region of the ith network interface region.
6. The quality identification method based on the router network line interface according to claim 1, wherein the obtaining the corrected enhancement index of each pixel point in each network line interface area based on the initial enhancement index of each sub-area of each network line interface area, the sharpness index of each network line interface area, and the initial enhancement index of each sub-area of the reference network line interface area includes:
for any pixel point in any network interface area:
acquiring a pixel point at the same position as the pixel point in the reference network interface area, and marking the pixel point as a matching point of the pixel point; calculating the absolute value of the difference value between the initial enhancement index of the subarea where the pixel point is located and the initial enhancement index of the subarea where the matching point of the pixel point is located, and recording the absolute value as a first absolute value; the sum of the natural constant 1 and the first absolute value is recorded as a second index;
calculating the ratio of the definition of the network line interface area where the pixel point is located to the definition of the network line interface area where the matching point of the pixel point is located, and marking the ratio as a third ratio; the difference value between the natural constant 1 and the third ratio is recorded as a third index;
and calculating the product of the initial enhancement index, the second index and the third index of the subarea where the pixel point is located, and taking the product as a correction enhancement index of the pixel point.
7. The quality recognition method based on the router network cable interface according to claim 1, wherein the obtaining the importance level of each gray level in each network cable interface region based on the correction enhancement index of each pixel point in each network cable interface region includes:
for any of the network interface areas:
and counting the gray levels in the network line interface area and the correction enhancement indexes of all the pixel points corresponding to the gray levels, and calculating the accumulated sum of the correction enhancement indexes of all the pixel points corresponding to the gray levels in the network line interface area as the importance degree of the corresponding gray levels in the network line interface area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311666486.8A CN117635922A (en) | 2023-12-06 | 2023-12-06 | Quality identification method based on router network cable interface |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311666486.8A CN117635922A (en) | 2023-12-06 | 2023-12-06 | Quality identification method based on router network cable interface |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117635922A true CN117635922A (en) | 2024-03-01 |
Family
ID=90016130
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311666486.8A Pending CN117635922A (en) | 2023-12-06 | 2023-12-06 | Quality identification method based on router network cable interface |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117635922A (en) |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| NL7401861A (en) * | 1973-02-12 | 1974-08-14 | ||
| US20160322006A1 (en) * | 2015-04-29 | 2016-11-03 | Shenzhen China Star Optoelectronics Technology Co. Ltd. | Dynamic backlight adjustment method of display screen |
| CN109564495A (en) * | 2016-06-28 | 2019-04-02 | 株式会社尼康 | Display device, program, display methods and control device |
| US20220076635A1 (en) * | 2020-09-04 | 2022-03-10 | Samsung Display Co., Ltd. | Display device and method of operating a display device |
| CN114359087A (en) * | 2021-12-29 | 2022-04-15 | 青岛海信移动通信技术股份有限公司 | Image processing method, device, equipment and medium |
| CN115330682A (en) * | 2022-07-11 | 2022-11-11 | 李雪利 | Chemical fiber surface stiff silk defect detection method based on texture analysis |
| CN115393657A (en) * | 2022-10-26 | 2022-11-25 | 金成技术股份有限公司 | Metal pipe production abnormity identification method based on image processing |
| CN115691026A (en) * | 2022-12-29 | 2023-02-03 | 湖北省林业科学研究院 | Intelligent early warning monitoring management method for forest fire prevention |
| CN115830459A (en) * | 2023-02-14 | 2023-03-21 | 山东省国土空间生态修复中心(山东省地质灾害防治技术指导中心、山东省土地储备中心) | Method for detecting damage degree of mountain forest and grass life community based on neural network |
| CN115861135A (en) * | 2023-03-01 | 2023-03-28 | 铜牛能源科技(山东)有限公司 | Image enhancement and identification method applied to box panoramic detection |
| CN115965624A (en) * | 2023-03-16 | 2023-04-14 | 山东宇驰新材料科技有限公司 | Detection method for anti-wear hydraulic oil pollution particles |
| CN116091499A (en) * | 2023-04-07 | 2023-05-09 | 山东中胜涂料有限公司 | Abnormal paint production identification system |
| CN116416166A (en) * | 2023-06-12 | 2023-07-11 | 贵州省人民医院 | Liver biopsy data analysis method and system |
| CN116739970A (en) * | 2022-03-08 | 2023-09-12 | 爱思开矽得荣株式会社 | Apparatus for evaluating edge defects in silicon wafer and method thereof |
-
2023
- 2023-12-06 CN CN202311666486.8A patent/CN117635922A/en active Pending
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| NL7401861A (en) * | 1973-02-12 | 1974-08-14 | ||
| US20160322006A1 (en) * | 2015-04-29 | 2016-11-03 | Shenzhen China Star Optoelectronics Technology Co. Ltd. | Dynamic backlight adjustment method of display screen |
| CN109564495A (en) * | 2016-06-28 | 2019-04-02 | 株式会社尼康 | Display device, program, display methods and control device |
| US20220076635A1 (en) * | 2020-09-04 | 2022-03-10 | Samsung Display Co., Ltd. | Display device and method of operating a display device |
| CN114359087A (en) * | 2021-12-29 | 2022-04-15 | 青岛海信移动通信技术股份有限公司 | Image processing method, device, equipment and medium |
| CN116739970A (en) * | 2022-03-08 | 2023-09-12 | 爱思开矽得荣株式会社 | Apparatus for evaluating edge defects in silicon wafer and method thereof |
| CN115330682A (en) * | 2022-07-11 | 2022-11-11 | 李雪利 | Chemical fiber surface stiff silk defect detection method based on texture analysis |
| CN115393657A (en) * | 2022-10-26 | 2022-11-25 | 金成技术股份有限公司 | Metal pipe production abnormity identification method based on image processing |
| CN115691026A (en) * | 2022-12-29 | 2023-02-03 | 湖北省林业科学研究院 | Intelligent early warning monitoring management method for forest fire prevention |
| CN115830459A (en) * | 2023-02-14 | 2023-03-21 | 山东省国土空间生态修复中心(山东省地质灾害防治技术指导中心、山东省土地储备中心) | Method for detecting damage degree of mountain forest and grass life community based on neural network |
| CN115861135A (en) * | 2023-03-01 | 2023-03-28 | 铜牛能源科技(山东)有限公司 | Image enhancement and identification method applied to box panoramic detection |
| CN115965624A (en) * | 2023-03-16 | 2023-04-14 | 山东宇驰新材料科技有限公司 | Detection method for anti-wear hydraulic oil pollution particles |
| CN116091499A (en) * | 2023-04-07 | 2023-05-09 | 山东中胜涂料有限公司 | Abnormal paint production identification system |
| CN116416166A (en) * | 2023-06-12 | 2023-07-11 | 贵州省人民医院 | Liver biopsy data analysis method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111161243B (en) | Industrial product surface defect detection method based on sample enhancement | |
| CN116912261B (en) | Plastic mold injection molding surface defect detection method | |
| CN110930390B (en) | Chip pin missing detection method based on semi-supervised deep learning | |
| CN104598886B (en) | A kind of method that utilization near-infrared high spectrum image recognizes the oil crops that go mouldy | |
| CN114723705A (en) | Cloth flaw detection method based on image processing | |
| CN115294116B (en) | Method, device and system for evaluating dyeing quality of textile material based on artificial intelligence | |
| CN115239718B (en) | Plastic product defect detection method and system based on image processing | |
| CN115170567B (en) | Method for detecting defects of waterproof steel plate for ship | |
| CN116758074B (en) | Multispectral food image intelligent enhancement method | |
| CN115953409B (en) | Injection molding surface defect detection method based on image processing | |
| CN116843581B (en) | Image enhancement method, system, device and storage medium for multi-scene graph | |
| CN114581376A (en) | Automatic sorting method and system for textile silkworm cocoons based on image recognition | |
| CN116563312A (en) | Method for dividing display image of double-screen machine | |
| CN117935296A (en) | Employment quality report generation system | |
| TWI498830B (en) | A method and system for license plate recognition under non-uniform illumination | |
| CN117635922A (en) | Quality identification method based on router network cable interface | |
| CN117274405A (en) | LED lamp working color detection method based on machine vision | |
| CN117197064A (en) | Automatic non-contact eye red degree analysis method | |
| CN116433568A (en) | LED lamp strip defect detection method and system based on depth support vector description | |
| CN117372420B (en) | Textile printing quality visual detection method in dyeing and finishing process | |
| CN111563536A (en) | Bamboo strip color self-adaptive classification method based on machine learning | |
| CN116645368B (en) | Online visual detection method for edge curl of casting film | |
| CN118447020B (en) | Printing typesetting abnormality detection method and device based on image recognition | |
| CN117115151B (en) | SIM card seat defect identification method based on machine vision | |
| CN117635595B (en) | Visual detection method for surface quality of precision injection mold embryo |
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 |