WO2021208423A1 - 一种基于pca学习的显卡接口机器测试方法 - Google Patents
一种基于pca学习的显卡接口机器测试方法 Download PDFInfo
- Publication number
- WO2021208423A1 WO2021208423A1 PCT/CN2020/129653 CN2020129653W WO2021208423A1 WO 2021208423 A1 WO2021208423 A1 WO 2021208423A1 CN 2020129653 W CN2020129653 W CN 2020129653W WO 2021208423 A1 WO2021208423 A1 WO 2021208423A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- card interface
- information
- graphics card
- output
- Prior art date
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N17/00—Diagnosis, testing or measuring for television systems or their details
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Definitions
- the invention belongs to the field of graphics card interface testing, and in particular relates to an automatic testing method of a graphics card interface.
- the present invention uses an industrial host computer equipped with a composite interface video collector to replace the display, collects the output signal of the graphics card interface through the video collector, and decodes it in the host computer, and realizes the automatic classification of video quality in combination with a machine learning method.
- the present invention provides an automated test method for a graphics card interface, which includes the following steps:
- Step 1 Connect the product to be tested to the video capture device with the corresponding model cable
- Step 2 Play the test video, the video collector collects the output information of the display interface, and uploads it to the industrial host computer;
- Step 3 The industrial host computer decodes the graphics card interface information and saves the frame image
- Step 4 Perform principal component analysis on the frame image to extract main information of the image
- each frame of image can be digitized into an m ⁇ n matrix A.
- Each element in matrix A represents a pixel in the picture.
- the frame image has a total of mn pixels.
- the size of m and n is determined by the test video. The resolution is determined.
- U and V are the characteristic matrices of AA T and A T A, respectively, and ⁇ is the following form:
- each frame of image A can be expanded into the following outer product form:
- Step 5 Use the analyzed image as the input of the machine learning classifier to obtain the detection result.
- the input layer of the neural network is composed of mn neurons, that is, mn pixels of the corresponding image A, and the hidden layer is composed of p neurons, and its output is
- the output layer is composed of 2 neurons, which represent qualified and unqualified respectively. Its expression is
- v j is the offset of the corresponding output
- v jh is the corresponding input t h to the output the weight of.
- step 2 retains the first k terms in the outer product form to extract the main information of the image
- step 3 the weights ⁇ hi , v jh and the bias ⁇ h , v j are obtained by training by solving the following optimization problem
- the advantages of the present invention are mainly embodied in that it can be compatible with various types of interfaces, does not need to be equipped with a display corresponding to the interface, uses a video collector and an industrial upper computer to complete the collection and decoding of interface information, and reduces the material cost.
- detection it combines principal component analysis and machine learning methods to achieve rapid and efficient automatic video quality detection, reducing labor costs and improving efficiency.
- Fig. 1 is a flowchart of an automated testing method for a graphics card interface of the present invention.
- Figure 2 is a framework diagram of an automated testing method for a graphics card interface of the present invention.
- Figure 3 is the simulation result of step four of the present invention.
- Figure 4 shows the detection accuracy of the present invention under different degrees of dimensionality reduction.
- Step 1 Connect the product to be tested to the industrial host computer with the video capture card installed with the corresponding type of wire;
- Step 2 Play the test video, and the video capture card collects the graphics interface information of the product to be tested and sends it to the industrial host computer
- Step 3 The industrial host computer decodes the graphics card interface information to generate and save a frame image
- Step 4 Perform principal component analysis on the frame image, the purpose of which is to extract the main information of the image.
- each frame of image can be digitized into an m ⁇ n matrix A.
- Each element in matrix A represents a pixel in the picture.
- the frame image has a total of mn pixels.
- the size of m and n is determined by the test video. The resolution is determined.
- U and V are the characteristic matrices of AA T and A T A, respectively, and ⁇ is the following form:
- each frame of image A can be expanded into the following outer product form:
- Step 5 Use the analyzed image as the input of the machine learning classifier to obtain the detection result.
- y l 1 indicates that the quality of A kl is unqualified
- the input layer of the neural network is composed of mn neurons, that is, mn pixels of the corresponding image A, and the hidden layer is composed of p neurons, and its output is
- the output layer is composed of 2 neurons, which represent qualified and unqualified respectively. Its expression is
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
Abstract
本发明公开了一种基于PCA学习的显卡接口机器测试方法,包括以下步骤:步骤1:用对应型号线材将待测产品连接到视频采集器;步骤2:播放测试视频,视频采集器收集显示接口输出信息,并上传至工业上位机;步骤3:工业上位机对显卡接口信息进行解码,并保存帧图像;步骤4:对所述帧图像进行主成分分析,提取图像主要信息;步骤5:将上述分析后的图像作为机器学习分类器的输入,从而得到检测结果。本发明能够兼容各种类型的接口,无需配备对应接口的显示器,利用视频采集器和工业上位机完成接口信息的采集与解码,降低了物力成本。
Description
本发明属于显卡接口测试领域,具体涉及一种显卡接口的自动化测试方法。
在显卡接口测试领域中,通常需要将待测产品接到具备对应类型接口的显示器,然后播放测试视频,由人工来检查视频显示质量是否合格。不仅在设备上需要配备各种接口类型的显示器,人工检查也存在因人工失误和检测时间长而影响生产效率,需要大量的人力物力成本。为此,本发明采用配备复合接口视频采集器的工业上位机替代显示器,通过视频采集器收集显卡接口输出信号,并在上位机中进行解码,结合机器学习方法实现视频质量的自动分类。
发明内容
为克服上述困难,本发明提供一种显卡接口的自动化测试方法,包括以下步骤:
步骤1:用对应型号线材将待测产品连接到视频采集器;
步骤2:播放测试视频,视频采集器收集显示接口输出信息,并上传至工业上位机;
步骤3:工业上位机对显卡接口信息进行解码,并保存帧图像;
步骤4:对所述帧图像进行主成分分析,提取图像主要信息;
首先每一帧图像都可以被数字化为一个m×n的矩阵A,矩阵A中的每一个元素都表示图片中的一个像素,帧图像一共有mn个像素,m和n的大小是由测试视频的分辨率决定的。对图像A进行奇异值分解,可得:
A=U∑V
T
其中U和V分别为AA
T和A
TA的特征矩阵,∑则为如下形式:
其中,σ
1>σ
2>…>σ
n>0为A的奇异值,因此,每一帧图像A都可以展开为如 下外积形式:
其中,u
1…u
n和v
1…v
n分别依次对应矩阵U和矩阵V的列向量。
步骤5:将上述分析后的图像作为机器学习分类器的输入,从而得到检测结果。其中,选择三层神经网络作为机器学习分类器的训练模型,训练数据
则由标记好的历史的检测数据得到,其中y
l=1表示A
kl质量不合格,y
l=0则表示A
kl质量合格。
神经网络的输入层由mn个神经元组成,即对应图像A的mn个像素,隐藏层由p个神经元组成,其输出为
最后,将待测待测产品的显卡接口输出图像作为输入,得到最终的检测结果。
优选的,步骤2保留外积形式的前k项从而提取图像主要信息
即保留原图像95%的信息。
优选的,步骤3中:最后通过求解以下最优化问题来训练得到权重ω
hi,v
jh和偏置ω
h,v
j
其中y
lj为样本A
kl的标签。
本发明的优势主要体现在:能够兼容各种类型的接口,无需配备对应接口的显示器,利用视频采集器和工业上位机完成接口信息的采集与解码,降低了物力成本。在检测方面,结合了主成分分析与机器学习方法,实现了快速高效的视频质量自动化检测,降低了人力成本,同时提高了效率。
图1为本发明一种显卡接口的自动化测试方法流程图。
图2为本发明一种显卡接口的自动化测试方法框架图。
图3为本发明步骤四的仿真结果。
图4为本发明降维不同程度下的检测精度。
下面结合附图1,2对本发明做进一步说明。
步骤1:用相应类型线材将待测产品连接安装了视频采集卡的工业上位机;
步骤2:播放测试视频,视频采集卡采集待测产品的显卡接口信息并发送到所述工业上位机
步骤3:所述工业上位机对所述显卡接口信息解码生成并保存帧图像;
步骤4:对所述帧图像进行主成分分析,其目的是提取图像主要信息。首先每一帧图像都可以被数字化为一个m×n的矩阵A,矩阵A中的每一个元素都表示图片中的一个像素,帧图像一共有mn个像素,m和n的大小是由测试视频的分辨率决定的。对图像A进行奇异值分解,可得:
A=U∑V
T
其中U和V分别为AA
T和A
TA的特征矩阵,∑则为如下形式:
其中,σ
1>σ
2>…>σ
n>0为A的奇异值,因此,每一帧图像A都可以展开为如下外积形式:
其中,u
1…u
n和v
1…v
n分别依次对应矩阵U和矩阵V的列向量,保留外积形式的前k项以实现提取图像主要信息的目的
即保留原图像95%的信息,其降维效果如图3所示;
步骤5:将上述分析后的图像作为机器学习分类器的输入,从而得到检测结果。其中,选择三层神经网络作为机器学习分类器的训练模型,训练数据
则由标记好的历史的检测数据得到,其中y
l=1表示A
kl质量不合格,y
l=0则表示A
kl质量合格。这里我们选择了。
神经网络的输入层由mn个神经元组成,即对应图像A的mn个像素,隐藏层由p个神经元组成,其输出为
其中y
lj为样本A
kl的标签。
我们选取了过往50个合格产品和50个不合格产品的输出信号,将其PCA降维处理后用于神经网络训练,得到训练好的神经网络分类器。最后,将待测待测产品的显卡接口输出图像作为输入,机器学习分类器输出合格或者不合格的分类结果。选取另外的50个合格产品和50个不合格产品做为测试集,用于检验训练结果,图4显示了不同降维程度下的检测精度,可以发现在提取特征维度在10到25之间时,检测准确率达到了99%,当特征过多或者过少时,检测精度均有所下降。最后,在应用中选取10个特征维度作为降维后的输入,完成100个产品的检测仅需2.3秒,实现了很高的检测效率。
如上所述,便可较好地实现本发明,上述实施例仅为本发明的典型实施例,并非用来限定本发明的实施范围,即凡依本发明内容所作的均等变化与修饰,都为本发明权利要求所要求保护的范围所涵盖。
Claims (4)
- 一种基于PCA学习的显卡接口机器测试方法,其特征在于:步骤1:用对应型号线材将待测产品连接到视频采集器;步骤2:播放测试视频,视频采集器收集显示接口输出信息,并上传至工业上位机;步骤3:工业上位机对显卡接口信息进行解码,并保存帧图像;步骤4:对所述帧图像进行主成分分析,提取图像主要信息;首先每一帧图像都可以被数字化为一个m×n的矩阵A,矩阵A中的每一个元素都表示图片中的一个像素,帧图像一共有mn个像素,m和n的大小是由测试视频的分辨率决定的;对图像A进行奇异值分解,可得:A=UΣV T其中U和V分别为AA T和A TA的特征矩阵,Σ则为如下形式:其中,σ 1>σ 2>…>σ n>0为A的奇异值,因此,每一帧图像A都可以展开为如下外积形式:其中,u 1…u n和v 1…v n分别依次对应矩阵U和矩阵V的列向量;步骤5:将上述分析后的图像作为机器学习分类器的输入,从而得到检测结果;其中,选择三层神经网络作为机器学习分类器的训练模型,训练数据S=(A kl,y l)则由标记好的历史的检测数据得到,其中y l=1表示A kl质量不合格,y l=0则表示A kl质量合格;神经网络的输入层由mn个神经元组成,即对应图像A的mn个像素,隐藏层由p个神经元组成,其输出为最后,将待测待测产品的显卡接口输出图像作为输入,得到最终的检测结果。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010290127.7A CN111526357B (zh) | 2020-04-14 | 2020-04-14 | 一种基于pca学习的显卡接口机器测试方法 |
CN202010290127.7 | 2020-04-14 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2021208423A1 true WO2021208423A1 (zh) | 2021-10-21 |
Family
ID=71902683
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2020/129653 WO2021208423A1 (zh) | 2020-04-14 | 2020-11-18 | 一种基于pca学习的显卡接口机器测试方法 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN111526357B (zh) |
WO (1) | WO2021208423A1 (zh) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117033253A (zh) * | 2023-10-10 | 2023-11-10 | 北京轻松怡康信息技术有限公司 | 一种接口测试方法、装置、电子设备及存储介质 |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111526357B (zh) * | 2020-04-14 | 2021-06-29 | 艾瑞思检测技术(苏州)有限公司 | 一种基于pca学习的显卡接口机器测试方法 |
CN113852809B (zh) * | 2021-09-27 | 2023-03-24 | 深圳市湘凡科技有限公司 | 一种音视频转发器件的测试系统及测试方法 |
CN114095795A (zh) * | 2021-11-26 | 2022-02-25 | 上海立磐软件科技有限公司 | 移动终端的视频质量检测方法 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103941112A (zh) * | 2013-01-18 | 2014-07-23 | 技嘉科技股份有限公司 | 多影像讯号的检测系统及其检测方法 |
CN106681868A (zh) * | 2015-11-10 | 2017-05-17 | 技嘉科技股份有限公司 | 图像数据测试方法及系统 |
CN109598695A (zh) * | 2017-09-29 | 2019-04-09 | 南京大学 | 一种基于深度学习网络的无参考图像模糊程度估计方法 |
US20190364268A1 (en) * | 2012-12-28 | 2019-11-28 | Promptlink Communications, Inc. | Video Quality Analysis and Detection of Blockiness, Artifacts and Color Variation for High-Volume Testing of Devices Using Automated Video Testing System |
CN111526357A (zh) * | 2020-04-14 | 2020-08-11 | 艾瑞思检测技术(苏州)有限公司 | 一种基于pca学习的显卡接口机器测试方法 |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7219085B2 (en) * | 2003-12-09 | 2007-05-15 | Microsoft Corporation | System and method for accelerating and optimizing the processing of machine learning techniques using a graphics processing unit |
CN101515243A (zh) * | 2008-02-21 | 2009-08-26 | 鸿富锦精密工业(深圳)有限公司 | 图形显示卡测试系统及方法 |
KR20100046460A (ko) * | 2008-10-27 | 2010-05-07 | 윤정문 | 컴퓨터용 고장진단장치 |
CN201765586U (zh) * | 2010-08-11 | 2011-03-16 | 名硕电脑(苏州)有限公司 | 显示卡测试系统 |
-
2020
- 2020-04-14 CN CN202010290127.7A patent/CN111526357B/zh active Active
- 2020-11-18 WO PCT/CN2020/129653 patent/WO2021208423A1/zh active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20190364268A1 (en) * | 2012-12-28 | 2019-11-28 | Promptlink Communications, Inc. | Video Quality Analysis and Detection of Blockiness, Artifacts and Color Variation for High-Volume Testing of Devices Using Automated Video Testing System |
CN103941112A (zh) * | 2013-01-18 | 2014-07-23 | 技嘉科技股份有限公司 | 多影像讯号的检测系统及其检测方法 |
CN106681868A (zh) * | 2015-11-10 | 2017-05-17 | 技嘉科技股份有限公司 | 图像数据测试方法及系统 |
CN109598695A (zh) * | 2017-09-29 | 2019-04-09 | 南京大学 | 一种基于深度学习网络的无参考图像模糊程度估计方法 |
CN111526357A (zh) * | 2020-04-14 | 2020-08-11 | 艾瑞思检测技术(苏州)有限公司 | 一种基于pca学习的显卡接口机器测试方法 |
Non-Patent Citations (2)
Title |
---|
"MATLAB Mathematical Modeling Methods and Practice, 3rd Edition", 31 July 2018, BEIJING UNIVERSITY OF AERONAUTICS AND ASTRONAUTICS PRESS, CN, ISBN: 9787512427273, article ZHUO, JINWU: "3.5 Micro data shackles - 3.5. 1 Basic Principles of Principal Component Analysis (PCA)", pages: 41 - 43, XP009531290 * |
"Vision-based unmarked gesture recognition", 30 June 2016, JILIN UNIVERSITY PRESS, CN, ISBN: 978-7-5677-6761-4, article ZHANG, SHENGJUN: "Passage; Vision-based unmarked gesture recognition", pages: 55 - 56, XP009531083 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117033253A (zh) * | 2023-10-10 | 2023-11-10 | 北京轻松怡康信息技术有限公司 | 一种接口测试方法、装置、电子设备及存储介质 |
Also Published As
Publication number | Publication date |
---|---|
CN111526357A (zh) | 2020-08-11 |
CN111526357B (zh) | 2021-06-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2021208423A1 (zh) | 一种基于pca学习的显卡接口机器测试方法 | |
CN111784633B (zh) | 一种面向电力巡检视频的绝缘子缺损自动检测算法 | |
WO2021208424A1 (zh) | 一种基于拉普拉斯特征映射学习的显卡接口机器测试方法 | |
CN111242185B (zh) | 一种基于深度学习的缺陷快速初筛方法及系统 | |
CN112464846B (zh) | 一种车站货运列车车厢异常故障的自动识别方法 | |
Puno et al. | Quality Assessment of Mangoes using Convolutional Neural Network | |
CN113077444A (zh) | 一种基于cnn的超声无损检测图像缺陷分类方法 | |
CN106645180A (zh) | 检查基板玻璃缺陷的方法、现场终端及服务器 | |
CN114973032A (zh) | 一种基于深度卷积神经网络的光伏板热斑检测方法及装置 | |
CN113780423A (zh) | 一种基于多尺度融合的单阶段目标检测神经网络及工业品表面缺陷检测模型 | |
CN115272777A (zh) | 面向输电场景的半监督图像解析方法 | |
CN111680738A (zh) | 一种用于苹果品质检测的筛选装置及其检测方法 | |
CN115147347A (zh) | 一种面向边缘计算的玛钢管件表面缺陷检测的方法 | |
CN112560668B (zh) | 一种基于场景先验知识的人体行为识别方法 | |
CN114596244A (zh) | 基于视觉处理和多特征融合的红外图像识别方法及系统 | |
CN116958073A (zh) | 基于注意力特征金字塔机制的小样本钢材缺陷检测方法 | |
CN116626050A (zh) | 一种基于图像识别的木材分类系统 | |
CN116485766A (zh) | 一种基于改进yolox的粮食不完善粒检测和计数方法 | |
CN113947567B (zh) | 基于多任务学习的缺陷检测方法 | |
CN114842295A (zh) | 绝缘子故障检测模型的获得方法、装置及电子设备 | |
TW202238527A (zh) | 基於頻譜影像之空氣汙染檢測方法 | |
CN112001396A (zh) | 一种轴承表面形变及文字混合缺陷图像检测方法 | |
CN111724352B (zh) | 一种基于核密度估计的贴片led瑕疵标注方法 | |
Zhang et al. | Real-time wheat unsound kernel classification detection based on improved YOLOv5 | |
WO2024108901A1 (zh) | 一种基于多光谱图像的电力设备区域检测方法及系统 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 20930929 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 20930929 Country of ref document: EP Kind code of ref document: A1 |