CN116486092A - Electromagnetic probe calibration piece identification method based on improved Hu invariant moment - Google Patents
Electromagnetic probe calibration piece identification method based on improved Hu invariant moment Download PDFInfo
- Publication number
- CN116486092A CN116486092A CN202310374477.5A CN202310374477A CN116486092A CN 116486092 A CN116486092 A CN 116486092A CN 202310374477 A CN202310374477 A CN 202310374477A CN 116486092 A CN116486092 A CN 116486092A
- Authority
- CN
- China
- Prior art keywords
- image
- moment
- improved
- identified
- color
- 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
- 239000000523 sample Substances 0.000 title claims abstract description 26
- 238000000034 method Methods 0.000 title claims abstract description 22
- 238000001914 filtration Methods 0.000 claims abstract description 10
- 239000013598 vector Substances 0.000 claims abstract description 10
- 239000003086 colorant Substances 0.000 claims abstract description 7
- 238000000691 measurement method Methods 0.000 claims abstract description 4
- 238000007781 pre-processing Methods 0.000 claims abstract description 4
- 230000015572 biosynthetic process Effects 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000013139 quantization Methods 0.000 claims description 3
- 238000003786 synthesis reaction Methods 0.000 claims description 3
- 238000005315 distribution function Methods 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000013519 translation Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000005070 sampling Methods 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
- G06V10/30—Noise filtering
-
- 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
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/30—Assessment of water resources
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention provides an electromagnetic probe calibration piece identification method based on improved Hu invariant moment, which comprises the following implementation steps: 1) Image preprocessing is carried out on the image to be identified and the template image, and the smooth image removes image noise by Gaussian filtering; 2) Converting RGB color vectors of an RGB image to be identified into a 72HSV histogram, extracting characteristic colors of a calibration piece and setting a high threshold and a low threshold; 3) Downsampling the selected area image by using a Gaussian pyramid to obtain a low-resolution image with clear target contour features; 4) Calculating to obtain 4 improved Hu invariant moments by utilizing the second-order and third-order center distances of the low-resolution images in the third step; 5) And matching the Hu moment of the template image and the image to be identified which is screened by the characteristic color by using a similarity measurement method, so as to realize the positioning of the electromagnetic probe calibration piece. The improved Hu invariant moment algorithm provided by the invention has the advantages of high identification precision, rapid identification, strong anti-interference capability and the like.
Description
Technical Field
The invention relates to the field of target identification and target detection, in particular to an electromagnetic probe calibration piece identification method based on improved Hu invariant moment.
Background
With the rapid development of electronic technology and products, the problem of electromagnetic interference in the field of integrated circuits is becoming more serious. Therefore, it is important to accurately detect and analyze electromagnetic interference by using the electromagnetic interference detection device. One of the most important components of electromagnetic interference detection and diagnosis devices is an electromagnetic probe, and the calibration of the electromagnetic probe directly determines the accuracy of the probe and has a direct influence on the final measurement result. Calibration of electromagnetic probes using microstrip line calibration elements is the most widely used method. Most of the calibration at present adopts a manual calibration mode, but the manual identification is low in efficiency, low in accuracy and poor in reliability, and is unfavorable for subsequent work. The obvious contour features of the microstrip line calibration piece are utilized to extract the Hu invariant moment and match with the template image, the Hu invariant moment has scale, translation and rotation invariance, the microstrip line calibration piece can be well identified and positioned, and the identification accuracy and the calibration efficiency can be improved.
However, the conventional Hu invariant moment has the problems of overlarge invariant moment calculated amount, long matching time, low object recognition precision for similar contours and the like.
Disclosure of Invention
In order to solve the problems, the invention discloses an electromagnetic probe calibration piece identification method based on improved Hu invariant moment, which is used for identifying and positioning a microstrip line calibration piece in an image according to Hu moment. The operation amount of Hu moment extraction and matching is reduced.
In order to achieve the above purpose, the invention provides an electromagnetic probe calibration piece identification method based on improved Hu invariant moment, which comprises the following specific steps:
step S1: image preprocessing is carried out on the image to be identified and the template image, and the smooth image removes image noise by Gaussian filtering;
step S2: converting RGB color vectors of an RGB image to be identified into a 72HSV histogram, extracting characteristic colors of a calibration piece, setting a high-low threshold value, and extracting an area in a characteristic color gamut range in the image to be identified;
step S3: downsampling the selected area image by using a Gaussian pyramid to obtain a low-resolution image with clear target contour features;
step S4: calculating to obtain 4 improved Hu invariant moments by utilizing the second-order and third-order center distances of the low-resolution images in the third step;
step S5: and matching the Hu moment of the template image and the image to be identified which is screened by the characteristic color by using a similarity measurement method, so as to realize the positioning of the electromagnetic probe calibration piece.
The further improvement is that: the step 1 specifically comprises the following steps:
step 11: filtering an image to be identified and a template image by using a two-dimensional Gaussian function H (x, y, sigma), removing noise points, wherein the expression of the two-dimensional Gaussian function is as follows:
the further improvement is that: the step 2 specifically comprises the following steps:
step 21 converts the R, G, B three color channel information of the RGB image to be identified into H, S, V of the HSV color space. The specific quantization conversion formula is as follows:
V＝max(R,G,B)
where R, G, B is the color intensity of the three color channel.
Step 22: converting the HSV three-dimensional vector into a one-dimensional 72HSV histogram, wherein the synthesis formula of the one-dimensional vector is as follows:
l＝9H+3S+3V
step 23: selecting characteristic color according to vivid color characteristics of the electromagnetic probe calibration, converting into 72HSV histogram by the method in step 21, 22, and setting high threshold value l based on the characteristic color l And a low threshold l h And selecting the color in the high and low threshold value interval as a color feature domain.
Step 24: traversing the image to be identified converted into 72HSV, and reserving pixel points in the color feature pair.
The further improvement is that: the step 3 specifically comprises the following steps:
step 31: using image pyramid theory, using Gaussian pyramid to downsample the image, and each downsampling layer to image the upper layer with Gaussian kernel G x Convolving to achieve Gaussian low-pass filtering, wherein the Gaussian convolution kernel G x The method comprises the following steps:
step 32: all even rows and even columns are deleted, so that the resolution of the image of the layer is reduced to 1/4 of that of the image of the upper layer.
The step 4 specifically comprises the following steps:
step 41: calculating normalized center distance by using geometric moment and center moment of the image:
wherein the geometric moment formula of the p+q order is as follows;
where f (x, y) is a density distribution function.
The p+q order central moment is:
wherein x is 0 ＝m 10 /m 00 ,y 0 ＝m 01 /m 00
The normalized center moment is:
step 42: calculating improved 4 Hu invariant moments of the image to be identified and the template image subjected to image processing by using the normalized center moment,
the step 5 specifically comprises the following steps:
and calculating errors by utilizing Euclidean distance, comparing each Hu moment characteristic of the identification image with Hu moment of the template image, setting a threshold value T, and taking a contour area with the error smaller than T as an identification success area. The formula of the Euclidean distance D (A, B) is:
the invention has the advantages and beneficial effects that:
1) The method utilizes the one-dimensional HSV histogram to pre-screen the target area of the image to be identified, eliminates the pixel areas with similar outlines and larger color characteristic differences, and improves the identification precision.
2) The invention uses the image pyramid to downwards sample the original image, reduces the resolution, thereby reducing the subsequent Hu moment operand
3) According to the normalized central moment, on the premise of keeping translation and rotation invariance, the conventional seven-dimensional Hu invariant moment is improved to 4-dimensional invariant moment, so that the matching operation amount is further reduced, and the matching speed is improved.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a microstrip line calibration piece picture;
FIG. 3 is a diagram of a color feature filtered outline region;
FIG. 4 is an image pyramid schematic;
FIG. 5 is an improved Hu moment value of the contour to be identified and the template image obtained in the example.
Detailed Description
The present invention is further illustrated in the following drawings and detailed description, which are to be understood as being merely illustrative of the invention and not limiting the scope of the invention. It should be noted that the words "front", "rear", "left", "right", "upper" and "lower" used in the following description refer to directions in the drawings, and the words "inner" and "outer" refer to directions toward or away from, respectively, the geometric center of a particular component.
An electromagnetic probe calibration piece identification method based on improved Hu invariant moment of the embodiment, a specific flow chart is shown in fig. 1, an image of the calibration piece is shown in fig. 2, and the method comprises the following steps:
step S1: image preprocessing is carried out on the image to be identified and the template image, and the smooth image removes image noise by Gaussian filtering;
step S2: converting RGB color vectors of an RGB image to be identified into a 72HSV histogram, extracting characteristic colors of a calibration piece, setting a high-low threshold value, and extracting an area in a characteristic color gamut range in the image to be identified; the screened regions are shown in FIG. 3.
Step S3: downsampling the selected region image by using a Gaussian pyramid to obtain a low-resolution image with clear target contour features, wherein the specific image pyramid sampling flow is shown in fig. 4;
step S4: calculating to obtain 4 improved Hu invariant moments by utilizing the second-order and third-order center distances of the low-resolution images in the third step;
step S5: and matching the Hu moment of the template image and the image to be identified which is screened by the characteristic color by using a similarity measurement method, so as to realize the positioning of the electromagnetic probe calibration piece.
The step 1 specifically comprises the following steps:
step 1: filtering an image to be identified and a template image by using a two-dimensional Gaussian function H (x, y, sigma), removing noise points, wherein the expression of the two-dimensional Gaussian function is as follows:
where x, y is the image pixel coordinates and σ is the gaussian distribution constant.
The step 2 specifically comprises the following steps:
step 2.1 converts the R, G, B three color channel information of the RGB image to be identified into H, S, V of the HSV color space. The specific quantization conversion formula is as follows:
V＝max(R,G,B)
step 2.2: converting the HSV three-dimensional vector into a one-dimensional 72HSV histogram, wherein the synthesis formula of the one-dimensional vector is as follows:
l＝9H+3S+3V
step 2.3: selecting characteristic color according to vivid color characteristics of an electromagnetic probe calibration piece, converting the characteristic color into a 72HSV histogram by using the methods in the steps 2.1 and 2.2, and setting a high threshold value l on the basis of the numerical value of the characteristic color l And a low threshold l h And selecting the color in the high and low threshold value interval as a color feature domain.
Step 3.4: traversing the image to be identified converted into 72HSV, and reserving pixel points in the color feature pair.
The step 3 specifically comprises the following steps:
step 3.1: using image pyramid theory, the image is downsampled by using Gaussian pyramid, the specific flow is as in FIG. 3, and each downsampling layer is to combine the upper image with Gaussian kernel G x Convolving to achieve Gaussian low-pass filtering, wherein the Gaussian convolution kernel G x The method comprises the following steps:
step 3.2: all even rows and even columns are deleted, so that the resolution of the image of the layer is reduced to 1/4 of that of the image of the upper layer.
The step 4 specifically comprises the following steps:
step 4.1: calculating normalized center distance by using geometric moment and center moment of the image:
wherein the geometric moment formula of the p+q order is as follows;
the p+q order central moment is:
the normalized center moment is:
step 4.2: calculating improved 4 Hu invariant moments of the image to be identified and the template image subjected to image processing by using the normalized center moment,
the step 5 specifically comprises the following steps:
each Hu moment feature of the identification image is compared with the Hu moment of the template image by using Euclidean distance to calculate. The formula of the Euclidean distance D (A, B) is:
in step 2, using the 72HSV histogram, the interference of the heterochromatic noise can be removed according to the color feature information of the calibration piece, and the regions with similar outlines and larger color feature differences are effectively screened. In step 3, the image pyramid is used in the improvement algorithm to sample the high-resolution 2506×1600 image of the industrial camera downwards, reduce the resolution thereof, obtain 320×200 image, and reduce the calculation amount of the improved Hu moment.
In step 5, the improved Hu moment pairs of the template calibration piece and the image calibration piece to be identified are shown in fig. 5, wherein it can be seen that the matching degree of the Hu moment of the template calibration piece and the image calibration piece to be identified is higher, and the improved Hu moment has the rotation, translation and scale invariance of invariant moment. And the calculation amount is small. Therefore, the calculation efficiency is higher, and when more objects exist in the background environment, the real-time performance of the optimized Hu moment algorithm is better.
The technical means disclosed by the scheme of the invention is not limited to the technical means disclosed by the embodiment, and also comprises the technical scheme formed by any combination of the technical features.
Claims (6)
1. An electromagnetic probe calibration identification method based on improved Hu invariant moment is characterized by comprising the following steps:
step S1: image preprocessing is carried out on the image to be identified and the template image, and the smooth image removes image noise by Gaussian filtering;
step S2: converting RGB color vectors of an RGB image to be identified into a 72HSV histogram, extracting characteristic colors of a calibration piece, setting a high-low threshold value, and extracting an area in a characteristic color gamut range in the image to be identified;
step S3: downsampling the selected area image by using a Gaussian pyramid to obtain a low-resolution image with clear target contour features;
step S4: calculating to obtain 4 improved Hu invariant moments by utilizing the second-order and third-order center distances of the low-resolution images in the third step;
step S5: and matching the Hu moment of the template image and the image to be identified which is screened by the characteristic color by using a similarity measurement method, so as to realize the positioning of the electromagnetic probe calibration piece.
2. The method for identifying the electromagnetic probe calibration based on the improved Hu invariant moment of claim 1, wherein: the step S1 specifically includes the following steps:
step 11: filtering an image to be identified and a template image by using a two-dimensional Gaussian function H (x, y, sigma), removing noise points, wherein the expression of the two-dimensional Gaussian function is as follows:
where x, y is the image pixel coordinates and σ is the gaussian distribution constant.
3. The method for identifying the electromagnetic probe calibration based on the improved Hu invariant moment of claim 1, wherein: the step S2 specifically includes the following steps:
step 21, converting R, G, B three-color channel information of the RGB image to be identified into H, S, V of HSV color space; the specific quantization conversion formula is as follows:
V＝max(R,G,B)
where R, G, B is the color intensity of the three color channel.
Step 22: converting the HSV three-dimensional vector into a one-dimensional 72HSV histogram, wherein the synthesis formula of the one-dimensional vector is as follows:
l＝9H+3S+3V
step 23: selecting characteristic colors according to vivid color features of the electromagnetic probe calibration, converting the characteristic colors into 72HSV histograms by using the methods in the steps 21 and 22, and setting a high threshold value ll and a low threshold value l on the basis of the numerical values of the characteristic colors h And selecting the color in the high and low threshold value interval as a color feature domain.
Step 24: traversing the image to be identified converted into 72HSV, and reserving pixel points in the color feature pair.
4. The method for identifying the electromagnetic probe calibration based on the improved Hu invariant moment of claim 1, wherein: the step S3 specifically includes the following steps:
step 31: using image pyramid theory, using Gaussian pyramid to downsample the image, and each downsampling layer to image the upper layer with Gaussian kernel G x The convolution is performed and the data is then processed,to achieve Gaussian low-pass filtering, where the Gaussian convolution kernel G x The method comprises the following steps:
step 32: all even rows and even columns are deleted, so that the resolution of the image of the layer is reduced to 1/4 of that of the image of the upper layer.
5. The method for identifying the electromagnetic probe calibration based on the improved Hu invariant moment of claim 1, wherein: the step 4 specifically comprises the following steps:
step 41: calculating normalized center distance by using geometric moment and center moment of the image:
wherein the geometric moment formula of the p+q order is as follows;
where f (x, y) is a density distribution function.
The p+q order central moment is:
wherein x is 0 ＝m 10 /m 00 ,y 0 ＝m 01 /m 00
The normalized center moment is:
step 42: calculating improved 4 Hu invariant moments of the image to be identified and the template image subjected to image processing by using the normalized center moment,
6. the method for identifying the electromagnetic probe calibration based on the improved Hu invariant moment of claim 1, wherein: the step 5 specifically comprises the following steps:
calculating by utilizing Euclidean distance, and comparing each Hu moment characteristic of the identification image with the Hu moment of the template image; the formula of the Euclidean distance D (A, B) is:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310374477.5A CN116486092A (en) | 2023-04-10 | 2023-04-10 | Electromagnetic probe calibration piece identification method based on improved Hu invariant moment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310374477.5A CN116486092A (en) | 2023-04-10 | 2023-04-10 | Electromagnetic probe calibration piece identification method based on improved Hu invariant moment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116486092A true CN116486092A (en) | 2023-07-25 |
Family
ID=87217016
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310374477.5A Pending CN116486092A (en) | 2023-04-10 | 2023-04-10 | Electromagnetic probe calibration piece identification method based on improved Hu invariant moment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116486092A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117607771A (en) * | 2023-10-09 | 2024-02-27 | 广东工业大学 | Electromagnetic signal measurement calibration system and multiport matrix transformation calibration method |
-
2023
- 2023-04-10 CN CN202310374477.5A patent/CN116486092A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117607771A (en) * | 2023-10-09 | 2024-02-27 | 广东工业大学 | Electromagnetic signal measurement calibration system and multiport matrix transformation calibration method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109344701B (en) | Kinect-based dynamic gesture recognition method | |
| CN110866903B (en) | Ping-pong ball identification method based on Hough circle transformation technology | |
| CN111915704A (en) | Apple hierarchical identification method based on deep learning | |
| CN107452030B (en) | Image registration method based on contour detection and feature matching | |
| Krishnan et al. | A survey on different edge detection techniques for image segmentation | |
| CN110222661B (en) | Feature extraction method for moving target identification and tracking | |
| CN116486092A (en) | Electromagnetic probe calibration piece identification method based on improved Hu invariant moment | |
| CN112861654A (en) | Famous tea picking point position information acquisition method based on machine vision | |
| CN113822810A (en) | Method for positioning workpiece in three-dimensional space based on machine vision | |
| CN111126296A (en) | Fruit positioning method and device | |
| CN113723314A (en) | Sugarcane stem node identification method based on YOLOv3 algorithm | |
| CN115661110B (en) | Transparent workpiece identification and positioning method | |
| CN114972453B (en) | Improved SAR image region registration method based on LSD and template matching | |
| CN113420633B (en) | Traffic sign identification method based on UM enhancement and SIFT feature extraction | |
| CN111274858A (en) | Business license identification method in network transaction supervision | |
| CN114998624A (en) | Image searching method and device | |
| CN111127311B (en) | Image registration method based on micro-coincident region | |
| CN109934215B (en) | Identification card identification method | |
| CN112541471A (en) | Shielded target identification method based on multi-feature fusion | |
| CN112329796A (en) | Infrared imaging cirrus cloud detection method and device based on visual saliency | |
| CN110956095A (en) | Multi-scale face detection method based on corner skin color detection | |
| Gupta et al. | A comparison of visual attention models for the salient text content detection in natural scene | |
| CN112070098B (en) | Hyperspectral image salient target detection method based on frequency adjustment model | |
| CN109376730B (en) | Gesture recognition method and device | |
| Satonkar | A Novel Approach Based Classification For Object Identification Using Digital Image Processing |
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 |