A method for detecting luminous intensity and uniformity based on CCD perception

Through the detection method based on CCD perception, the brightness and uniformity characteristics of the luminous surface are extracted and the learning model is established, which solves the problems of inefficiency and noise interference of traditional detection methods, and achieves efficient and accurate detection of luminous intensity and uniformity.

CN112767527BActive Publication Date: 2024-11-22WUHAN HAIWEI TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110118040.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-01-28
Publication Date
2024-11-22
Estimated Expiration
2041-01-28

AI Technical Summary

Technical Problem

Traditional dot luminance meters cannot meet the efficient detection needs of modern display devices, especially inefficient in full-screen detection and different luminous area detection, and are susceptible to noise interference, so they cannot adapt to the specific detection needs of different products.

Method used

Using a detection method based on CCD perception, by imaging the luminous surface to be measured in a dark room environment, extracting the brightness and uniformity feature matrix, establishing and learning the brightness model, combining the multi-dimensional feature matrix and uniformity matrix to calculate the luminous intensity and uniformity, using CCD performs inspections to improve detection accuracy and efficiency.

Benefits of technology

It realizes detection of different luminous materials and locations, resists noise interference, improves detection accuracy and efficiency, simplifies the calibration process, reduces costs, and is suitable for mass production.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112767527B_ABST
    Figure CN112767527B_ABST
Patent Text Reader

Abstract

The present invention relates to a detection method for luminous intensity and uniformity based on CCD perception, including a calibration step and a detection step, wherein the detection step includes the following steps: step a1, taking a luminous surface to be measured, setting the parameters of the CCD to be the same as the CCD parameters of the calibration step; step a2, imaging the luminous surface to be measured in a darkroom environment to obtain an imaging image I; step a3, extracting the brightness feature matrix M and uniformity feature U of the imaging image I; step a4, establishing and learning a brightness model according to the calibration step to obtain a learned brightness model; step a5, calculating the brightness A of the luminous surface to be measured according to the learned brightness model and the brightness feature matrix; and calculating the uniformity u of the luminous surface to be measured according to the uniformity feature matrix U. The present invention can realize the brightness and uniformity detection of different luminous materials, and can realize the detection of different luminous positions and different luminous sizes, can resist noise interference, and is simple and easy to implement.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of luminous surface detection, and in particular to a detection method for luminous intensity and uniformity based on CCD perception. Background Art

[0002] With the rapid upgrading of flat panel displays, users have higher and higher requirements for the display effect of display devices, as well as high requirements for production efficiency in mass production, making it impossible for traditional point brightness meters to meet the test requirements. Traditional brightness and uniformity detection methods are mainly 5-point and 9-point methods, which results in extremely low efficiency when full-screen detection is required. Whether it is an automatic or semi-automatic production line, it takes a long time and cannot meet the needs of the production line.

[0003] In addition, traditional methods are not good for detecting luminous brightness in a small range. Some methods have a complicated calibration process. For different products, the calibration method is single and cannot adapt well to different display devices. Some devices cannot measure the luminous brightness at any point. In the automated detection of display devices, the brightness change of display devices shows a certain trend. Traditional algorithms cannot fit this trend well for specific products and cannot achieve accurate detection of specific products. Some detection equipment is easily affected by noise.

[0004] Therefore, it will be necessary to have testing equipment that can perform full-screen testing, testing of different luminous areas, reduce noise interference, and adapt to different production lines and different products. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a method for detecting luminous intensity and uniformity based on CCD perception in view of the above shortcomings.

[0006] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0007] A method for detecting luminous intensity and uniformity based on CCD perception comprises the following steps:

[0008] Step a1, take the luminous surface to be measured, and set the CCD parameters to be the same as the CCD parameters in the calibration step;

[0009] Step a2: in a darkroom environment, imaging the luminous surface to be measured to obtain an imaging image I;

[0010] Step a3, extracting the brightness feature matrix M and uniformity feature U of the imaged image I;

[0011] Step a4, establishing and learning a brightness model according to the calibration step to obtain a learned brightness model;

[0012] Step a5: Calculate the brightness A of the light-emitting surface to be measured according to the learned brightness model and the brightness characteristic matrix; calculate the uniformity u of the light-emitting surface to be measured according to the uniformity characteristic matrix U.

[0013] Furthermore, the step a3 specifically includes the following steps:

[0014] Step a31, generating a k-layer pyramid G of the imaging image I, expressed as:

[0015] G={G1…G j}, j = 1, 2, ..., k;

[0016] Step a32: G j Divide into m*n regions of equal size, j=1, 2, ..., k;

[0017] Step a34, extract G in sequence j , j = 1, 2, ..., the gray value G at the same position on k j (p, q), (p, q) are the coordinates of the position point, p = 1, 2, ..., m; q = 1, 2, ..., n, forming a scale-consistent local histogram, expressed as:

[0018] H p,q (j) = G j (p, q); j=1, 2,..., k; p=1, 2,..., m; q=1, 2,..., n;

[0019] Step a35: All local histograms H with consistent sizes p,q (j) Merge to obtain the multi-dimensional feature matrix H of the image I, expressed as:

[0020] H={H p,q (j)|j=1,2,…,k;p=1,2,…,m;q=1,2,…,n}

[0021] Then, the brightness feature matrix M of the image I is expressed as:

[0022]

[0023] Step a36: normalize the multidimensional feature matrix H, expressed as:

[0024]

[0025] Generate the uniformity feature matrix U of the imaging image I, expressed as:

[0026]

[0027] Furthermore, the calibration step includes the following steps:

[0028] Step b1, selecting N luminous surface samples that are the same as the luminous surface to be measured, and setting the brightness values ​​of the N luminous surface samples according to the arithmetic transformation rule;

[0029] Step b2: Divide each luminous surface sample into m*n areas, and use a high-precision measurement tool to measure the brightness of each area of ​​the luminous surface sample. The brightness standard value of the N luminous surface samples is expressed as:

[0030]

[0031] Among them, A i is the brightness standard value of the i-th luminous surface sample, i = 1, 2...N;

[0032] Step b3, setting CCD parameters;

[0033] Step b4: In a dark room, place N luminous surface samples in sequence directly under the CCD and image them to generate N imaging images B. i , i = 1, 2, ..., N, and the corresponding imaging image sequence {B i}i=1,2,…,N;

[0034] Step b5: Extracting the image sequence i Each image B in i The brightness feature matrix Mi, the extraction steps are the same as step a3;

[0035] Step b6: Establish brightness model:

[0036]

[0037] Among them, M i Image B i The brightness feature matrix, A i is the brightness standard value of the i-th luminous surface sample, α, β and γ are the values ​​to be trained in the brightness model;

[0038] Step b7, all images B i Substitute the brightness feature matrix Mi into the brightness model, train the brightness model, and obtain the trained brightness model.

[0039] Furthermore, in step b4, the formula for calculating the uniformity of the light-emitting surface to be measured is:

[0040]

[0041] Where u is the uniformity of the light-emitting surface to be measured, and U is the uniformity matrix.

[0042] After adopting the above technical solution, the present invention has the following advantages compared with the prior art:

[0043] The present invention can realize the brightness and uniformity detection of different luminous materials, and can realize the detection of different luminous positions and different luminous sizes, can resist noise interference, and is simple and easy to implement; the detection method of the present invention establishes a brightness learning model through a calibration process, and selects multiple sample luminous surfaces to train the learning model, thereby improving the accuracy of the learning model value to improve the detection accuracy; the present invention uses CCD for inspection, has high detection efficiency, low cost and is easy to promote.

[0044] The present invention is described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A sample of the luminous surface (i.e., calibration sample) whose brightness is displayed according to an arithmetic progression to be consistent with the luminous surface to be tested;

[0046] Figure 2 Schematic diagram of the brightness matrix and uniformity covariance matrix obtained for constructing a Gaussian pyramid;

[0047] Figure 3 Schematic diagram of the imaging system. DETAILED DESCRIPTION

[0048] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0049] A method for detecting luminous intensity and uniformity based on CCD perception comprises the following steps:

[0050] Step a1, take the luminous surface to be measured, and set the parameters of CCD, including focal length f, aperture F, gain GA, white balance parameter BA, exposure time T and working distance L, so that they are the same as the GCD parameters in the calibration step;

[0051] Step a2: in a darkroom environment, imaging the luminous surface to be measured to obtain an imaging image I;

[0052] Step a3: extracting the brightness feature matrix M and uniformity feature U of the image I. The extraction process specifically includes the following steps:

[0053] Step a31: Figure 2 As shown, the k-layer pyramid G of the imaging image I is generated, which is expressed as:

[0054] G={G1…G j}, j = 1, 2, ..., k;

[0055] Step a32: Gj Divide into m*n regions of equal size, j=1, 2, ..., k;

[0056] Step a34, extract G in sequence j , j = 1, 2, ..., the gray value G at the same position on k j (p, q), (p, q) are the coordinates of the position point, p = 1, 2, ..., m; q = 1, 2, ..., n, forming a scale-consistent local histogram, expressed as:

[0057] H p,q (j) = G j (p, q); j=1, 2,..., k; p=1, 2,..., m; q=1, 2,..., n;

[0058] Step a35: All local histograms H with consistent sizes p,q (j) Merge to obtain the multi-dimensional feature matrix H of the image I, expressed as:

[0059] H={H p,q (j)|j=1,2,…,k;p=1,2,…,m;q=1,2,…,n}

[0060] Then, the brightness feature matrix M of the image I is expressed as:

[0061]

[0062] Step a36: normalize the multidimensional feature matrix H, expressed as:

[0063]

[0064] Generate the uniformity feature matrix U of the imaging image I, expressed as:

[0065]

[0066] Step a4, establishing and learning a brightness model according to the calibration step to obtain a learned brightness model;

[0067] Step a5: Calculate the brightness A of the light-emitting surface to be measured according to the learned brightness model and the brightness feature matrix, and substitute the brightness feature matrix M into the learned brightness model to obtain the brightness A of the light-emitting surface, which is expressed as:

[0068] A=αM 2 +βM+γ;

[0069] According to the uniformity characteristic matrix U, the uniformity u of the light-emitting surface to be measured is calculated. The formula for calculating the uniformity of the light-emitting surface to be measured is:

[0070]

[0071] Where u is the uniformity of the light-emitting surface to be measured, and U is the uniformity matrix.

[0072] The calibration procedure includes the following steps:

[0073] Step b1: Figure 1 As shown, N luminous surface samples that are the same as the luminous surface to be measured are selected, and the brightness values ​​of the N luminous surface samples are set according to the arithmetic transformation rule;

[0074] Step b2: Divide each luminous surface sample into m*n areas, and use a high-precision measurement tool to measure the brightness of each area of ​​the luminous surface sample. The brightness standard value of the N luminous surface samples is expressed as:

[0075]

[0076] Among them, A i is the brightness standard value of the i-th luminous surface sample, i = 1, 2...N;

[0077] Step b3, setting CCD parameters;

[0078] Step b4: In a dark room, place N luminous surface samples in sequence directly under the CCD and image them to generate N imaging images B. i , i = 1, 2, ..., N, and the corresponding imaging image sequence {B i}i=1,2,…,N;

[0079] Step b5: Extracting the image sequence i Each image B in i The brightness feature matrix Mi, the extraction steps are the same as step a3;

[0080] Step b6: Establish brightness model:

[0081]

[0082] Among them, M i Image B i The brightness feature matrix, A i is the brightness standard value of the i-th luminous surface sample, α, β and γ are the values ​​to be trained in the brightness model;

[0083] Step b7, all images B i Substitute the brightness feature matrix Mi into the brightness model, train the brightness model, and obtain the trained brightness model.

[0084] like Figure 3As shown, a luminous intensity and uniformity detection system based on CCD perception includes a CCD imaging device and a host computer. The CCD imaging device is mainly used to image the luminous surface to be detected, and the host computer is used to control imaging, brightness model learning, brightness detection and uniformity detection.

[0085] The above is an example of the best implementation of the present invention, and the parts not described in detail are common knowledge of ordinary technicians in the field. The protection scope of the present invention shall be based on the content of the claims, and any equivalent transformation based on the technical enlightenment of the present invention is also within the protection scope of the present invention.

Claims

1. A method for detecting luminous intensity and uniformity based on CCD perception, characterized in that: The following steps are involved: Step a1, take the luminous surface to be measured, and set the CCD parameters to be the same as the CCD parameters in the calibration step; Step a2: in a darkroom environment, imaging the luminous surface to be measured to obtain an imaging image I; Step a3, extracting the brightness feature matrix M and uniformity feature matrix U of the imaged image I; The step a3 specifically comprises the following steps: Step a31, generating a k-layer pyramid G of the imaging image I, expressed as: G={G1…G j },j=1,2,…,k; Step a32: G j Divide into m*n regions of equal size, j=1,2,…,k; Step a34, extract G in sequence j ,j=1,2,…,kThe gray value G at the same position j (p, q), (p, q) are the coordinates of the position point, p = 1, 2, ..., m; q = 1, 2, ..., n, forming a scale-consistent local histogram, expressed as: H p,q (j)=G j (p,q);j=1,2,…,k;p=1,2,…,m;q=1,2,…,n; Step a35: All local histograms H with consistent sizes p,q (j) Merge to obtain the multi-dimensional feature matrix H of the image I, expressed as: H={H p,q (j)|j=1,2,…,k;p=1,2,…,m;q=1,2,…,n} Then, the brightness feature matrix M of the image I is expressed as: Step a36: normalize the multidimensional feature matrix H, expressed as: Generate the uniformity feature matrix U of the imaging image I, expressed as: Step a4, establishing and learning a brightness model according to the calibration step to obtain a learned brightness model; The calibration step comprises the following steps: Step b1, selecting N luminous surface samples that are the same as the luminous surface to be measured, and setting the brightness values ​​of the N luminous surface samples according to the arithmetic transformation rule; Step b2: Divide each luminous surface sample into m*n areas, and use a high-precision measurement tool to measure the brightness of each area of ​​the luminous surface sample. The brightness standard value of the N luminous surface samples is expressed as: Among them, A i is the brightness standard value of the i-th luminous surface sample, i = 1, 2...N; Step b3, setting CCD parameters; Step b4: In a dark room, place N luminous surface samples in sequence directly under the CCD and image them to generate N imaging images B. i , i = 1, 2, ..., N, and the corresponding imaging image sequence {B i }i=1,2,…,N; Step b5: Extracting the image sequence i Each image B in i The brightness feature matrix Mi, the extraction steps are the same as step a3; Step b6: Establish brightness model: Among them, M i Image B i The brightness feature matrix, A i is the brightness standard value of the i-th luminous surface sample, α, β and γ are the values ​​to be trained in the brightness model; Step b7, all images B i Substitute the brightness feature matrix Mi into the brightness model, train the brightness model, and obtain the trained brightness model; Step a5: Calculate the brightness A of the light-emitting surface to be measured according to the learned brightness model and the brightness characteristic matrix; calculate the uniformity u of the light-emitting surface to be measured according to the uniformity characteristic matrix U.

2. The method for detecting luminous intensity and uniformity based on CCD perception according to claim 1, characterized in that: In step b4, the formula for calculating the uniformity of the light-emitting surface to be measured is: Where u is the uniformity of the light-emitting surface to be measured, and U is the uniformity characteristic matrix.