CN102721702A - Distributed paper defect detection system and method based on embedded processor - Google Patents
Distributed paper defect detection system and method based on embedded processor Download PDFInfo
- Publication number
- CN102721702A CN102721702A CN2012102155366A CN201210215536A CN102721702A CN 102721702 A CN102721702 A CN 102721702A CN 2012102155366 A CN2012102155366 A CN 2012102155366A CN 201210215536 A CN201210215536 A CN 201210215536A CN 102721702 A CN102721702 A CN 102721702A
- Authority
- CN
- China
- Prior art keywords
- paper
- image
- detection
- paper defect
- camera
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 157
- 238000001514 detection method Methods 0.000 title claims abstract description 127
- 238000000034 method Methods 0.000 title claims abstract description 11
- 230000037303 wrinkles Effects 0.000 claims abstract description 12
- 238000007781 pre-processing Methods 0.000 claims description 30
- 238000006243 chemical reaction Methods 0.000 claims description 14
- 238000001914 filtration Methods 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 5
- 206010027146 Melanoderma Diseases 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 4
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims description 3
- 201000010099 disease Diseases 0.000 claims description 2
- 230000006872 improvement Effects 0.000 abstract description 4
- 238000010276 construction Methods 0.000 abstract description 2
- 206010040954 Skin wrinkling Diseases 0.000 description 9
- 238000012545 processing Methods 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 3
- 238000012546 transfer Methods 0.000 description 3
- 208000000260 Warts Diseases 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000010191 image analysis Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 201000010153 skin papilloma Diseases 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
Images
Landscapes
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
本发明涉及一种基于嵌入式处理器的分布式纸病检测系统及方法，它采用并排的多路工业用线阵CCD相机采集运动纸幅图像数据，图像数据经CameraLink线缆传送到各自对应的嵌入式纸病检测处理器。嵌入式纸病检测处理器先利用FPGA对图像进行采集、预处理，再利用DSP检测图像中存在的纸病的类型、面积、位置等数据。各检测处理器的检测结果及对应的纸病图像通过千兆以太网传送至中心服务器。中心服务器通过子窗口对多路检测结果及纸病图像进行更新显示，并将其保存到后台数据库中。该系统具有构建成本低、速度提升空间大、易于实现分布式检测优点，可应用于车速不低于1km/min的运动纸幅的黑斑、孔洞、褶皱、刮痕等常见纸病高速在线检测。
The invention relates to a distributed paper defect detection system and method based on an embedded processor. It adopts multiple industrial line array CCD cameras arranged side by side to collect moving paper web image data, and the image data is transmitted to the respective corresponding Embedded web defect detection processor. The embedded paper defect detection processor first uses FPGA to collect and preprocess the image, and then uses DSP to detect the type, area, location and other data of the paper defect in the image. The detection results of each detection processor and the corresponding paper defect images are transmitted to the central server through Gigabit Ethernet. The central server updates and displays the multi-channel detection results and paper defect images through the sub-window, and saves them in the background database. The system has the advantages of low construction cost, large space for speed improvement, and easy implementation of distributed detection. It can be applied to high-speed online detection of common paper defects such as black spots, holes, wrinkles, and scratches on moving paper webs with a speed of no less than 1km/min. .
Description
the
技术领域 technical field
本发明涉及基于机器视觉的纸张缺陷检测技术领域，尤其涉及一种基于嵌入式处理器的分布式纸病检测系统及方法。 The invention relates to the technical field of paper defect detection based on machine vision, in particular to a distributed paper defect detection system and method based on an embedded processor.
背景技术 Background technique
纸张生产过程中，纸幅表面会不可避免地产生一些缺陷，即纸病，如：斑点、孔洞、皱褶、刮痕等,这些外观纸病是会严重影响纸张质量，因此纸病检测是生产过程中的一个重要环节。传统的纸病检测由人工完成，但现代造纸业具有纸幅宽、速度快的特点，依靠人工检测纸页纸病已不能满足要求。近年来，随着机器计算机和数字图像处理技术的发展，基于机器视觉的纸病在线监测系统在一些大型造纸企业得到了广泛应用，该系统能够快速、有效的实现纸幅纸病在线检测。但在这种系统中，纸病检测主要依靠运行在PC机上的图像处理软件实现，这种检测方案一方面不利于实现大型分布式在线检测，另一方面由于其构建成本一般都比较高，在中小规模的造纸企业的应用受到限制。针对这些问题，基于嵌入式处理器的机器视觉技术逐步应用到的纸病检测领域中，并出现了一些应用系统，这些系统大多采用高速DSP芯片作为纸病图像分析的核心处理器，一定程度上克服了基于PC机的纸病检测的缺点，但由于纸病图像的采集、预处理、纸病检测等所有的处理操作都由DSP完成， DSP的本身运算负荷就非常大，而且DSP芯片还需频繁与外部存储设备进行数据交换，因此对于实时在线检测，时间裕量不够，采用这种技术方案的速度提升有限。 In the process of paper production, some defects will inevitably occur on the surface of the paper web, that is, paper defects, such as: spots, holes, wrinkles, scratches, etc. These appearance paper defects will seriously affect the quality of paper, so paper defect detection is a production an important part of the process. The traditional detection of paper defects is done manually, but the modern paper industry has the characteristics of wide paper width and fast speed, and relying on manual detection of paper defects can no longer meet the requirements. In recent years, with the development of machine computers and digital image processing technology, the online detection system for paper defects based on machine vision has been widely used in some large paper-making enterprises. This system can quickly and effectively realize the online detection of paper defects. But in this kind of system, the detection of paper defects mainly depends on the image processing software running on the PC. On the one hand, this kind of detection scheme is not conducive to the realization of large-scale distributed online detection. The application of small and medium-sized papermaking enterprises is limited. In response to these problems, machine vision technology based on embedded processors has been gradually applied to the field of paper defect detection, and some application systems have emerged. Most of these systems use high-speed DSP chips as the core processor for paper defect image analysis. It overcomes the shortcomings of PC-based paper defect detection, but since all processing operations such as paper defect image collection, preprocessing, and paper defect detection are completed by DSP, the calculation load of DSP itself is very large, and the DSP chip also needs Frequent data exchange with external storage devices, so for real-time online detection, the time margin is not enough, and the speed improvement of this technical solution is limited.
发明内容 Contents of the invention
本发明的目的就是为了克服现有纸病检测技术存在的缺陷，提供一种基于嵌入式处理器的分布式纸病检测系统，其构成成本低、速度提升空间大、易于实现分布式检测，可实现黑斑、孔洞、褶皱、刮痕等常见纸病的在线检测。 The purpose of the present invention is to overcome the defects existing in the existing paper defect detection technology, and provide a distributed paper defect detection system based on an embedded processor, which has low construction cost, large speed improvement space, and is easy to realize distributed detection. Realize online detection of common paper defects such as black spots, holes, folds, scratches, etc. the
为了实现上述目的，本发明采用如下技术方案： In order to achieve the above object, the present invention adopts the following technical solutions:
一种基于嵌入式处理器的分布式纸病检测系统，它包括至少一组高速相机，高速相机安装在运动纸幅上方，在运动纸幅下方则设有LED光源；高速相机与嵌入式纸病检测处理器连接，嵌入式纸病检测处理器通过以太网交换机与中心服务器通信；其中，嵌入式纸病检测处理器由FPGA和DSP协作实现，包括FPGA内的图像采集单元，它与FPGA内的双口RAM、纸病图像预处理单元、以太网接口控制单元以及FPGA外的千兆网控制与收发芯片依次通信；纸病图像预处理单元与纸病检测DSP、高速SRAM通信；图像采集单元通过LVDS/ LVTTL转换芯片采集图像数据，通过LVTTL/ LVDS转换芯片控制高速相机；纸病图像预处理单元接收图像采集单元的图像数据，并对其进行预处理，纸病检测DSP接收纸病图像预处理单元的图像数据，并进行纸病检测；以太网接口控制单元一方面控制千兆网控制与收发芯片接收来自以太网的相机设置参数，并将其传送至图像采集单元，另一方面接收来自图像预处理单元的纸病检测结果数据，并控制千兆网控制与收发芯片将其传送至以太网。 A distributed paper defect detection system based on an embedded processor, which includes at least one group of high-speed cameras, the high-speed camera is installed above the moving paper web, and an LED light source is provided below the moving paper web; the high-speed camera and the embedded paper defect The detection processor is connected, and the embedded paper defect detection processor communicates with the central server through an Ethernet switch; among them, the embedded paper defect detection processor is realized by cooperation of FPGA and DSP, including the image acquisition unit in FPGA, which communicates with the Dual-port RAM, web defect image preprocessing unit, Ethernet interface control unit, and gigabit network control outside the FPGA communicate with the transceiver chip in sequence; the web defect image preprocessing unit communicates with the web defect detection DSP and high-speed SRAM; the image acquisition unit passes The LVDS/LVTTL conversion chip collects image data, and controls the high-speed camera through the LVTTL/LVDS conversion chip; the paper defect image preprocessing unit receives the image data of the image acquisition unit and preprocesses it, and the paper defect detection DSP receives the paper defect image preprocessing The image data of the unit, and to detect paper defects; on the one hand, the Ethernet interface control unit controls the Gigabit network control and transceiver chip to receive the camera setting parameters from the Ethernet, and transmit them to the image acquisition unit; The paper defect detection result data of the preprocessing unit is controlled and transmitted to the Ethernet by controlling the Gigabit network control and transceiver chip.
所述高速相机由多路工业用线阵CCD相机组成，横向并排安装在运动纸幅的上方；每路相机通过Camera Link线缆连接到各自的嵌入式纸病检测处理器。 The high-speed camera is composed of multiple industrial linear array CCD cameras, which are installed side by side above the moving paper web; each camera is connected to its own embedded paper defect detection processor through a Camera Link cable.
一种采用基于嵌入式处理器的分布式纸病检测系统的检测方法，它的步骤为： A detection method using a distributed paper defect detection system based on an embedded processor, its steps are:
（1）参数设置 (1) Parameter setting
通过中心服务器的前台检测软件操作，对各路高速相机的行频、曝光时间以及图像尺寸等参数进行设置； Through the front-end detection software operation of the central server, set the line frequency, exposure time and image size and other parameters of each high-speed camera;
（2）检测启动 (2) Detection start
参数设置完成后，通过中心服务器的前台检测软件发出纸病检测命令，该命令通过以太网传送至各路嵌入式纸病检测处理器，启动在线纸病检测； After the parameter setting is completed, the front-end detection software of the central server sends a paper defect detection command, and the command is transmitted to each embedded paper defect detection processor through Ethernet to start the online paper defect detection;
（3）纸页图像采集 (3) Paper page image acquisition
各路嵌入式纸病检测处理器通过FPGA内的图像采集单元和Camera Link接口将相机采集的纸页图像数据保存到FPGA内的双口RAM中； Each embedded paper defect detection processor saves the paper image data collected by the camera to the dual-port RAM in the FPGA through the image acquisition unit in the FPGA and the Camera Link interface;
（4）图像预处理 (4) Image preprocessing
嵌入式纸病检测处理器通过FPGA内的图像预处理单元从双口RAM中读取图像数据，并对其进行中值滤波及分段灰度线性变换预处理； The embedded paper defect detection processor reads the image data from the dual-port RAM through the image preprocessing unit in the FPGA, and performs median filtering and segmented gray scale linear transformation preprocessing on it;
（5）纸病检测与定位 (5) Paper defect detection and positioning
预处理后的图像数据输入纸病检测DSP，纸病检测DSP利用阈值分割法将图像中亮斑、刮痕类高亮度纸病区域和黑斑、褶皱类低亮度纸病区域分别从背景中分离出来；然后计算各个纸病区域的圆形度，并根据圆形度区分孔洞与刮痕、黑斑与褶皱； The preprocessed image data is input into the paper defect detection DSP, and the paper defect detection DSP uses the threshold segmentation method to separate the bright spot, scratch-like high-brightness paper defect area and black spot, wrinkle-like low-brightness paper defect area in the image from the background. Then calculate the circularity of each paper defect area, and distinguish holes and scratches, black spots and wrinkles according to the circularity;
（6）如果纸病检测DSP检测到纸页图像中存在纸病，则将每个纸病的类型、面积、位置参数及相应纸病图像通过千兆以太网传送至中心服务器； (6) If the paper defect detection DSP detects that there is a paper defect in the paper image, the type, area, position parameter and corresponding paper defect image of each paper defect are transmitted to the central server through Gigabit Ethernet;
（7）中心服务器通过多个子窗口对多路检测结果及纸病图像进行更新显示，同时将其保存到后台数据库中。 (7) The central server updates and displays the multi-channel detection results and paper defect images through multiple sub-windows, and saves them in the background database at the same time.
所述步骤（1）中，在中心服务器上的检测软件将各路高速相机的行频、曝光时间、图像大小等参数从硬盘上直接调入内存，然后通过以太网传送至嵌入式纸病检测处理器中FPGA内的图像采集单元，图像采集单元由此产生相机控制信号，并由LVTTL/LVDS转换芯片和Camera Link线缆传送至高速相机。 In the above step (1), the detection software on the central server directly transfers the line frequency, exposure time, image size and other parameters of each high-speed camera into the memory from the hard disk, and then transmits them to the embedded paper defect detection via Ethernet The image acquisition unit in the FPGA in the processor, from which the image acquisition unit generates camera control signals, which are transmitted to the high-speed camera by the LVTTL/LVDS conversion chip and Camera Link cable.
所述步骤（3）中，高速相机输出图像的有效像素数据经Camera Link连接电缆和LVDS/LVTTL接口转换芯片，传送至FPGA内的图像采集单元；图像采集单元以行同步信号作为写有效信号，以像素时钟作为写时钟将高速相机的图像数据写入到双口RAM中。 In the step (3), the effective pixel data of the high-speed camera output image is transmitted to the image acquisition unit in the FPGA through the Camera Link connection cable and the LVDS/LVTTL interface conversion chip; the image acquisition unit uses the line synchronization signal as the write effective signal, Write the image data of the high-speed camera into the dual-port RAM with the pixel clock as the write clock.
所述步骤（4）中，当采集到一帧图像数据后，FPGA内的图像预处理单元读取双口RAM中的图像数据，先根据图像尺寸对图像进行边界裁剪处理，然后进行中值滤波，以消除随机噪声干扰，之后再进行分段线性灰度变换，增强图像中的纸病区域，抑制其背景区域；预处理结果保存到FPGA外的高速SRAM中；预处理结束后，图像预处理单元向纸病检测DSP发出中断信号。 In the step (4), when a frame of image data is collected, the image preprocessing unit in the FPGA reads the image data in the dual-port RAM, first performs boundary cropping processing on the image according to the image size, and then performs median filtering , to eliminate random noise interference, and then perform segmented linear grayscale transformation to enhance the paper defect area in the image and suppress its background area; the preprocessing results are saved to the high-speed SRAM outside the FPGA; after the preprocessing is completed, the image preprocessing The unit sends an interrupt signal to the paper defect detection DSP.
所述步骤（5）中，纸病检测DSP接收到中断信号后，读取SRAM中图像数据；先利用阈值分割出图像中孔洞和刮痕类高亮度纸病区域以及斑点、褶皱类低亮度纸病区域；利用开运算去除噪声干扰后，再利用标记法确定图像中各个纸病区域的位置及面积；然后计算各个纸病区域周长平方与面积比得到圆形度，圆形度较大的区域为褶皱或刮痕，圆形度较小区域的为黑斑与亮斑。 In the step (5), after the paper defect detection DSP receives the interrupt signal, it reads the image data in the SRAM; firstly, the threshold value is used to segment the high-brightness paper defect areas such as holes and scratches in the image, and the low-brightness paper such as spots and wrinkles disease area; use the open operation to remove noise interference, and then use the marking method to determine the position and area of each paper defect area in the image; then calculate the square of the perimeter of each paper defect area and the area ratio to obtain the circularity, and the circularity is larger Areas are creases or scratches, and areas of less circularity are dark and bright spots.
所述步骤（6）中，纸病检测DSP完成图像检测后，如果纸病的个数不为零，则将每个纸病的类型、面积、位置等参数传送至FPGA的以太网接口控制单元，该单元将纸病参数、纸病图像以及所属的相机号数据进行打包，并控制千兆网控制与收发芯片将数据传送至以太网，再由以太网传送至中心服务器。 In the step (6), after the paper defect detection DSP completes the image detection, if the number of paper defects is not zero, the parameters such as the type, area, and position of each paper defect are transmitted to the Ethernet interface control unit of the FPGA , the unit packs the data of paper wart parameters, paper wart images and the corresponding camera number, and controls the Gigabit network control and transceiver chip to transmit the data to the Ethernet, and then transmits the data to the central server through the Ethernet.
所述步骤（7）中，在中心服务器的前台检测软件界面上，每个相机都有一个对应的子窗口，用于显示该路相机检测结果；软件接收到以太网传送过来的数据后，根据其所属的相机号将采集的纸病图像及相应的纸病参数显示在对应的子窗口中，并将其保存到后台数据库中。 In the step (7), on the foreground detection software interface of the central server, each camera has a corresponding sub-window for displaying the detection result of the camera; after the software receives the data transmitted by the Ethernet, it The camera number to which it belongs will display the collected web defect image and the corresponding web defect parameters in the corresponding sub-window, and save it in the background database.
本发明采用了并排的多路工业用线阵CCD相机采集运动纸幅图像数据，图像数据经Camera Link接口传送至各自对应的嵌入式纸病检测处理器，由于Camera Link接口的传输速率最高可达2.1Gbps，可以保证数据采集的高速性。 The present invention adopts multi-channel industrial linear array CCD cameras arranged side by side to collect the image data of the moving paper web, and the image data are transmitted to the corresponding embedded paper defect detection processors through the Camera Link interface, because the transmission rate of the Camera Link interface can reach up to 2.1Gbps, can guarantee the high speed of data collection.
嵌入式纸病检测处理器通过对纸病图像进行分析处理实现纸病检测，是整个检测系统的核心，也是制约整个检测系统速度的瓶颈。为了保证图像分析处理的高速性，该单元由FPGA和DSP协作实现，充分利用DSP的高速数据处理能力和FPGA的复杂逻辑处理能力。嵌入式纸病检测处理器完成纸病检测后，检测结果及对应的纸病图像通过千兆以太网传送至中心服务器。中心服务器由计算机及运行在计算机上的前台检测软件及后台数据库组成。服务器接收各路检测单元的检测结果，并进行显示和保存。 The embedded web defect detection processor realizes the web defect detection by analyzing and processing the web defect image, which is the core of the whole detection system and also the bottleneck restricting the speed of the whole detection system. In order to ensure the high speed of image analysis and processing, this unit is realized by cooperation of FPGA and DSP, making full use of the high-speed data processing ability of DSP and the complex logic processing ability of FPGA. After the embedded paper defect detection processor completes the paper defect detection, the detection result and the corresponding paper defect image are transmitted to the central server through Gigabit Ethernet. The central server is composed of a computer, foreground detection software running on the computer and a background database. The server receives the detection results of each detection unit, and displays and saves them.
本发明的有益效果是：通过Camera Link接口线阵CCD相机实现了纸页图像的快速采集；利用FPGA和DSP嵌入式处理器实现了快速纸病检测与定位；利用千兆以太网保证了检测结果和纸病图像的快速传输。构建的纸病在线检测系统具有成本低、速度提升空间大、易于实现分布式检测优点，可应用于车速不低于1km/min的运动纸幅的黑斑、孔洞、褶皱、刮痕等常见纸病高速在线检测。 The beneficial effects of the present invention are: the rapid collection of paper image is realized through the linear array CCD camera with Camera Link interface; the rapid detection and positioning of paper defects is realized by using FPGA and DSP embedded processor; the detection result is guaranteed by using Gigabit Ethernet Fast transfer of paper and paper images. The constructed online detection system for paper defects has the advantages of low cost, large space for speed improvement, and easy implementation of distributed detection. It can be applied to common paper such as black spots, holes, wrinkles, and scratches on moving paper webs with a speed of not less than 1km/min. Disease high-speed online detection.
附图说明：Description of drawings:
图1基于嵌入式处理器的分布式纸病检测系统结构示意图； Figure 1 is a schematic structural diagram of a distributed paper defect detection system based on an embedded processor;
图2嵌入式纸病检测处理器原理框图； Fig. 2 Principle block diagram of embedded paper defect detection processor;
图3 纸病检测算法总体流程图； Figure 3 The overall flow chart of the paper defect detection algorithm;
图4 确定纸病面积、位置及分离孔洞、刮痕的算法流程图； Fig. 4 Algorithm flow chart for determining the area and position of the paper defect and separating holes and scratches;
图5 示例纸病图像及纸病类型识别结果。 Figure 5 is an example of paper defect images and recognition results of paper defect types.
其中，1.高速相机，2. 运动纸幅，3.LED光源，4.嵌入式纸病检测处理器，5. 以太网交换机，6.中心服务器，7. LVDS/ LVTTL转换芯片，8. LVTTL/LVDS转换芯片，9.图像采集单元，10.双口RAM，11.图像预处理单元，12.纸病检测DSP，13.高速SRAM，14.以太网接口控制单元，15.千兆网控制与收发芯片。 Among them, 1. High-speed camera, 2. Moving paper web, 3. LED light source, 4. Embedded paper detection processor, 5. Ethernet switch, 6. Central server, 7. LVDS/LVTTL conversion chip, 8. LVTTL /LVDS conversion chip, 9. Image acquisition unit, 10. Dual-port RAM, 11. Image preprocessing unit, 12. Paper defect detection DSP, 13. High-speed SRAM, 14. Ethernet interface control unit, 15. Gigabit network control and transceiver chip.
具体实施方式 Detailed ways
下面结合附图和实施例对本发明作进一步说明。 The present invention will be further described below in conjunction with drawings and embodiments.
如图1所示，基于嵌入式处理器的分布式纸病检测系统主要包含LED光源3、运动纸幅2、高速相机1、嵌入式纸病检测处理器4、以太网交换机5和中心服务器6。LED光源3安装在运动纸幅2的下方，高速相机1由多路线阵CCD相机组成，横向并排安装在运动纸幅2的上方。每路相机通过Camera Link接口连接到各自的嵌入式纸病检测处理器4。如图2所示，嵌入式纸病检测处理器4主要由FPGA内的图像采集单元9、纸病图像预处理单元11、以太网接口控制单元14、双口RAM10以及纸病检测DSP12组成，此外还包括片外高速SRAM13和LVDS/LVTTL转换芯片7、LVTTL/LVDS转换芯片8、千兆网控制与收发芯片15等外部接口芯片。各路嵌入式纸病检测处理器4内的以太网接口控制单元14通过以千兆网控制与收发芯片15连接到以太网交换机5，再由以太网交换机5连接到中心服务器6。中心服务器6由计算机及运行在计算机上的前台检测软件及后台数据库组成。 As shown in Figure 1, the distributed paper defect detection system based on embedded processor mainly includes LED light source 3, moving paper web 2, high-speed camera 1, embedded paper defect detection processor 4, Ethernet switch 5 and central server 6 . The LED light source 3 is installed under the moving paper web 2 , and the high-speed camera 1 is composed of multiple line array CCD cameras, and is installed side by side above the moving paper web 2 horizontally. Each camera is connected to its own embedded paper defect detection processor 4 through a Camera Link interface. As shown in Figure 2, the embedded paper defect detection processor 4 is mainly composed of an image acquisition unit 9 in the FPGA, a paper defect image preprocessing unit 11, an Ethernet interface control unit 14, a dual-port RAM10 and a paper defect detection DSP12. It also includes off-chip high-speed SRAM13, LVDS/LVTTL conversion chip 7, LVTTL/LVDS conversion chip 8, Gigabit network control and transceiver chip 15 and other external interface chips. The Ethernet interface control unit 14 in each embedded paper defect detection processor 4 is connected to the Ethernet switch 5 through the Gigabit network control and transceiver chip 15, and then connected to the central server 6 through the Ethernet switch 5. Central server 6 is made up of computer and the foreground detection software and background database running on the computer.
本发明的纸病检测方法为： Paper defect detection method of the present invention is:
（1）参数设置。中心服务器6的前台检测软件从硬盘上调入各路高速相机1的行频、曝光时间、图像大小等设置参数，设置参数通过以太网传送至嵌入式纸病检测处理器4。如图2所示，FPGA内的图像采集单元9产生TTL形式的相机控制信号，该控制信号由LVTTL/ LVDS转换芯片8转换成LVDS形式，并由Camera Link连接电缆传送至各路高速相机1。 (1) Parameter setting. The foreground detection software of the central server 6 transfers setting parameters such as the line frequency, exposure time, and image size of each high-speed camera 1 from the hard disk, and the setting parameters are transmitted to the embedded paper defect detection processor 4 through Ethernet. As shown in Figure 2, the image acquisition unit 9 in the FPGA generates a camera control signal in TTL form, which is converted into LVDS form by the LVTTL/LVDS conversion chip 8, and transmitted to each high-speed camera 1 by the Camera Link connecting cable.
（2）检测启动。通过软件界面发出纸病检测命令，该命令通过以太网传送至嵌入式纸病检测处理器4，启动在线纸病检测。 (2) Detection starts. A paper defect detection command is issued through the software interface, and the command is transmitted to the embedded paper defect detection processor 4 through the Ethernet to start online paper defect detection.
（3）图像数据采集。如图2所示，高速相机1输出的图像数据经Camera Link连接电缆传送至LVDS到 LVTTL转换芯片7，该芯片将4对LVDS数据信号和1对LVDS时钟信号转换成TTL 形式28位数据和1路时钟信号。28位数据中包含帧同步信号（FVAL，本系统中无效）、行有同步信号（LVAL）。FPGA内的图像采集单元9以LVAL作为写有效信号，以像素时钟作为写时钟将高速相机1的图像数据写入FPGA内的双口RAM10中。存储在双口RAM10中的是8位灰度图像，共256个灰度等级。 (3) Image data acquisition. As shown in Figure 2, the image data output by the high-speed camera 1 is transmitted to the LVDS to LVTTL conversion chip 7 through the Camera Link connecting cable, and the chip converts 4 pairs of LVDS data signals and 1 pair of LVDS clock signals into TTL form 28-bit data and 1 road clock signal. The 28-bit data includes frame synchronization signal (FVAL, invalid in this system) and line synchronization signal (LVAL). The image acquisition unit 9 in the FPGA writes the image data of the high-speed camera 1 into the dual-port RAM 10 in the FPGA with the LVAL as the write valid signal and the pixel clock as the write clock. What is stored in the dual-port RAM10 is an 8-bit grayscale image, with a total of 256 grayscale levels.
（4）图像预处理。当采集到一帧图像数据后，FPGA内的图像预处理单元9读取双口RAM10的图像数据，如图3所示，先根据图像尺寸对图像进行边界裁剪处理，再利用3×3的十字形模板进行中值滤波，然后利用分段线性灰度变换对纸病区域进行增强处理，图5a是一幅经过预处理后纸病图像（512×512）。预处理后的图像数据存入到高速SRAM13中。预处理结束后，FPGA内的图像预处理单元9向纸病检测DSP12发出中断请求。 (4) Image preprocessing. After collecting a frame of image data, the image preprocessing unit 9 in the FPGA reads the image data of the dual-port RAM 10, as shown in Figure 3, first performs boundary cutting processing on the image according to the image size, and then uses 3×3 ten The glyph template is subjected to median filtering, and then the piecewise linear grayscale transformation is used to enhance the paper defect area. Figure 5a is a preprocessed paper defect image (512×512). The preprocessed image data is stored in the high-speed SRAM13. After the preprocessing is finished, the image preprocessing unit 9 in the FPGA sends an interrupt request to the paper defect detection DSP12.
（5）纸病检测。纸病检测DSP12接收到中断请求后，读取高速SRAM13中图像数据，并进行纸病检测与定位。如图5a所示，由于LED光源3在运动纸幅2的下方，孔洞和刮痕属于图像中的高亮度区域，斑点及褶皱是属于图像中的低亮度区域。如图3所示，利用较小的经验阈值T 1分割出黑斑、褶皱等低亮度纸病区域，分割结果如图5b所示。再利用开运算去除噪声，标记连通域确定纸病位置、面积，然后计算各个区域的圆形度，圆形度较小的为黑斑，圆形度较大的为褶皱，如图4所示。黑斑识别结果如图5c所示，褶皱识别结果如图5d所示。如图3所示，利用较大的经验阈值T 2分割出图像中孔洞、刮痕等高亮度纸病区域，分割结果如图5e所示。再用同样的方法区分孔洞和刮痕并确定其位置、面积参数。孔洞识别结果如图5f所示，刮痕识别结果如图5g所示。 (5) Paper defect detection. Paper defect detection DSP12 reads the image data in high-speed SRAM13 after receiving the interrupt request, and performs paper defect detection and positioning. As shown in Fig. 5a, since the LED light source 3 is below the moving paper web 2, the holes and scratches belong to the high-brightness areas in the image, and the spots and wrinkles belong to the low-brightness areas in the image. As shown in Figure 3, a small empirical threshold T 1 is used to segment low-brightness paper defect areas such as black spots and wrinkles, and the segmentation results are shown in Figure 5b. Then use the open operation to remove noise, mark the connected domain to determine the location and area of the paper defect, and then calculate the circularity of each area. The smaller circularity is a black spot, and the larger circularity is a wrinkle, as shown in Figure 4 . The black spot recognition result is shown in Figure 5c, and the wrinkle recognition result is shown in Figure 5d. As shown in Figure 3, a large empirical threshold T 2 is used to segment high-brightness paper defect areas such as holes and scratches in the image, and the segmentation results are shown in Figure 5e. Then use the same method to distinguish holes and scratches and determine their position and area parameters. The hole recognition results are shown in Figure 5f, and the scratch recognition results are shown in Figure 5g.
（6）检测结果传送。纸病检测DSP12完成图像检测后，如果纸病的个数不为零，则将每个纸病的类型、面积、位置等数据传至FPGA内的以太网接口控制单元14，该单元将纸病参数、纸病图像以及所属的高速相机号进行打包，再通过千兆网控制与收发芯片15将数据传送至以太网，再由以太网传送至中心服务器6。 (6) Test result transmission. After the paper defect detection DSP12 completes the image detection, if the number of the paper defect is not zero, then the data such as the type, area, and position of each paper defect are transmitted to the Ethernet interface control unit 14 in the FPGA, and this unit will record the paper defect Parameters, paper defect images and the corresponding high-speed camera numbers are packaged, and then the data is transmitted to the Ethernet through the Gigabit network control and transceiver chip 15, and then transmitted to the central server 6 through the Ethernet.
（7）检测结果更新、显示。在中心服务器6的前台检测软件界面上，每路高速相机1都有一个对应的子窗口，用于显示该路高速相机1检测结果。软件接收到以太网传送过来的数据后，一方面根据其所属的高速相机号将纸病图像及相应的纸病参数显示在对应的子窗口中，另一方面将其保存到中心服务器6的后台数据库中。 (7) The test result is updated and displayed. On the foreground detection software interface of the central server 6, each high-speed camera 1 has a corresponding sub-window for displaying the detection result of the high-speed camera 1 . After the software receives the data sent by the Ethernet, on the one hand, it will display the paper defect image and the corresponding paper defect parameters in the corresponding sub-window according to the high-speed camera number to which it belongs, and on the other hand, it will save it to the background of the central server 6 in the database.
Claims (9)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012102155366A CN102721702A (en) | 2012-06-27 | 2012-06-27 | Distributed paper defect detection system and method based on embedded processor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012102155366A CN102721702A (en) | 2012-06-27 | 2012-06-27 | Distributed paper defect detection system and method based on embedded processor |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102721702A true CN102721702A (en) | 2012-10-10 |
Family
ID=46947524
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2012102155366A Pending CN102721702A (en) | 2012-06-27 | 2012-06-27 | Distributed paper defect detection system and method based on embedded processor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102721702A (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103399016A (en) * | 2013-07-26 | 2013-11-20 | 齐鲁工业大学 | Online detection system for surface defects of coldly-rolled aluminum plate and detection method of online detection system |
| CN103543161A (en) * | 2013-10-16 | 2014-01-29 | 湖南镭目科技有限公司 | On-line detection method of continuous casting billet surface quality |
| CN104458749A (en) * | 2013-09-25 | 2015-03-25 | 中国科学院沈阳自动化研究所 | Real-time aluminum profile surface defect detection system based on machine vision |
| CN104827766A (en) * | 2015-05-28 | 2015-08-12 | 安庆市德创机电产品设计有限公司 | Paper operation monitoring component of laminating machine |
| CN105203557A (en) * | 2015-10-28 | 2015-12-30 | 江西弘泰电子信息材料有限公司 | Electronic paper tape papermaking quality online detection system |
| CN107610316A (en) * | 2017-08-28 | 2018-01-19 | 深圳怡化电脑股份有限公司 | Detect the method, apparatus and terminal device of bank note defect |
| CN107976450A (en) * | 2016-10-21 | 2018-05-01 | 特克斯玛格销售有限公司 | Method and apparatus for material roll observation and material roll inspection |
| CN109342450A (en) * | 2018-10-30 | 2019-02-15 | 深圳灵图慧视科技有限公司 | Intelligent cloth examination device |
| CN110458223A (en) * | 2019-08-06 | 2019-11-15 | 湖南省华芯医疗器械有限公司 | Tumor of bronchus automatic testing method and detection system under a kind of scope |
| CN111519467A (en) * | 2020-03-20 | 2020-08-11 | 金东纸业（江苏）股份有限公司 | Coating head control method based on paper defect detection and related device |
| CN113504241A (en) * | 2021-07-13 | 2021-10-15 | 陕西科技大学 | Design method of paper defect detection system |
| CN115561252A (en) * | 2022-10-28 | 2023-01-03 | 四川轻化工大学 | High-speed ultra-wide paper defect intelligent detection system and method |
| CN115604406A (en) * | 2022-09-29 | 2023-01-13 | 南通伊诺精密塑胶导管有限公司（Cn） | Electronic endoscope image processor system based on self-adaptive adjustment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007333732A (en) * | 2006-05-18 | 2007-12-27 | Nippon Steel Corp | Surface inspection system and diagnostic method for inspection performance of surface inspection system |
| JP2008134107A (en) * | 2006-11-27 | 2008-06-12 | Daio Paper Corp | Detection method of flaw log and flaw log detector |
| CN202002894U (en) * | 2011-01-10 | 2011-10-05 | 山东轻工业学院 | Quick online paper flaw detecting system based on machine vision |
| CN202083645U (en) * | 2011-05-04 | 2011-12-21 | 浙江大学 | An FPGA-based on-line detection device for grain quality |
| CN202221415U (en) * | 2011-09-03 | 2012-05-16 | 山东轻工业学院 | Wide-width paper surface defect online detection system |
| CN202275060U (en) * | 2011-09-23 | 2012-06-13 | 山东轻工业学院 | Online paper defect inspection system |
| CN202676617U (en) * | 2012-06-27 | 2013-01-16 | 山东轻工业学院 | Embedded processor-based distributed paper defect detection system |
-
2012
- 2012-06-27 CN CN2012102155366A patent/CN102721702A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007333732A (en) * | 2006-05-18 | 2007-12-27 | Nippon Steel Corp | Surface inspection system and diagnostic method for inspection performance of surface inspection system |
| JP2008134107A (en) * | 2006-11-27 | 2008-06-12 | Daio Paper Corp | Detection method of flaw log and flaw log detector |
| CN202002894U (en) * | 2011-01-10 | 2011-10-05 | 山东轻工业学院 | Quick online paper flaw detecting system based on machine vision |
| CN202083645U (en) * | 2011-05-04 | 2011-12-21 | 浙江大学 | An FPGA-based on-line detection device for grain quality |
| CN202221415U (en) * | 2011-09-03 | 2012-05-16 | 山东轻工业学院 | Wide-width paper surface defect online detection system |
| CN202275060U (en) * | 2011-09-23 | 2012-06-13 | 山东轻工业学院 | Online paper defect inspection system |
| CN202676617U (en) * | 2012-06-27 | 2013-01-16 | 山东轻工业学院 | Embedded processor-based distributed paper defect detection system |
Non-Patent Citations (2)
| Title |
|---|
| 古帅，邱书波: "基于FPGA+DSP的纸张缺陷视觉检测系统的设计", 《中华纸业》, vol. 31, no. 6, 31 March 2010 (2010-03-31) * |
| 杨波等: "基于几何及灰度特征的纸病检测算法研究", 《中国造纸》, vol. 30, no. 9, 15 September 2011 (2011-09-15) * |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103399016A (en) * | 2013-07-26 | 2013-11-20 | 齐鲁工业大学 | Online detection system for surface defects of coldly-rolled aluminum plate and detection method of online detection system |
| CN104458749A (en) * | 2013-09-25 | 2015-03-25 | 中国科学院沈阳自动化研究所 | Real-time aluminum profile surface defect detection system based on machine vision |
| CN103543161A (en) * | 2013-10-16 | 2014-01-29 | 湖南镭目科技有限公司 | On-line detection method of continuous casting billet surface quality |
| CN104827766A (en) * | 2015-05-28 | 2015-08-12 | 安庆市德创机电产品设计有限公司 | Paper operation monitoring component of laminating machine |
| CN105203557A (en) * | 2015-10-28 | 2015-12-30 | 江西弘泰电子信息材料有限公司 | Electronic paper tape papermaking quality online detection system |
| CN105203557B (en) * | 2015-10-28 | 2018-06-08 | 江西洁美电子信息材料有限公司 | A kind of Electronic Paper band papermaking quality on-line detection system |
| CN107976450A (en) * | 2016-10-21 | 2018-05-01 | 特克斯玛格销售有限公司 | Method and apparatus for material roll observation and material roll inspection |
| CN107976450B (en) * | 2016-10-21 | 2021-10-29 | 特克斯玛格销售有限公司 | Method and apparatus for web observation and web inspection |
| CN107610316B (en) * | 2017-08-28 | 2019-12-10 | 深圳怡化电脑股份有限公司 | Method and device for detecting defect of paper money and terminal equipment |
| CN107610316A (en) * | 2017-08-28 | 2018-01-19 | 深圳怡化电脑股份有限公司 | Detect the method, apparatus and terminal device of bank note defect |
| CN109342450A (en) * | 2018-10-30 | 2019-02-15 | 深圳灵图慧视科技有限公司 | Intelligent cloth examination device |
| CN110458223A (en) * | 2019-08-06 | 2019-11-15 | 湖南省华芯医疗器械有限公司 | Tumor of bronchus automatic testing method and detection system under a kind of scope |
| CN110458223B (en) * | 2019-08-06 | 2023-03-17 | 湖南省华芯医疗器械有限公司 | Automatic detection method and detection system for bronchial tumor under endoscope |
| CN111519467A (en) * | 2020-03-20 | 2020-08-11 | 金东纸业（江苏）股份有限公司 | Coating head control method based on paper defect detection and related device |
| CN113504241A (en) * | 2021-07-13 | 2021-10-15 | 陕西科技大学 | Design method of paper defect detection system |
| CN115604406A (en) * | 2022-09-29 | 2023-01-13 | 南通伊诺精密塑胶导管有限公司（Cn） | Electronic endoscope image processor system based on self-adaptive adjustment |
| CN115561252A (en) * | 2022-10-28 | 2023-01-03 | 四川轻化工大学 | High-speed ultra-wide paper defect intelligent detection system and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102721702A (en) | Distributed paper defect detection system and method based on embedded processor | |
| CN111612763B (en) | Mobile phone screen defect detection method, device and system, computer equipment and medium | |
| JP7018135B2 (en) | Battery detection method and equipment | |
| CN108037138A (en) | A kind of web inspection system and detection method for being used to detect the two-sided defect of paper | |
| CN202221415U (en) | Wide-width paper surface defect online detection system | |
| CN103399016A (en) | Online detection system for surface defects of coldly-rolled aluminum plate and detection method of online detection system | |
| CN101510085B (en) | Fluoroplastic film defect on-line detecting control system based on process control machine | |
| CN100498820C (en) | Automatic recognising method and automatic recognising recording system for number of paper money | |
| CN107862692A (en) | A kind of ribbon mark of break defect inspection method based on convolutional neural networks | |
| CN205317684U (en) | Flaw detecting system dyes cloth based on image | |
| CN201512707U (en) | A printing error detection device based on an intelligent camera | |
| CN113751920B (en) | Embedded device and method for detecting welding quality of lockhole TIG welding in real time | |
| CN101726498B (en) | Intelligent detector and method of copper strip surface quality on basis of vision bionics | |
| CN105044126A (en) | Visual detection system for large-width continuous surface defects | |
| CN112132828A (en) | Film flaw detection method based on deep learning | |
| CN111915565A (en) | Method for analyzing cracks of porcelain insulator of power transmission and transformation line in real time based on YOLACT algorithm | |
| CN115587998A (en) | A method for detecting surface defects of metal sheets | |
| CN105825219A (en) | Machine vision detection system | |
| CN206891989U (en) | Paper defects data acquisition device and the real-time detection using the device and reponse system | |
| CN202676617U (en) | Embedded processor-based distributed paper defect detection system | |
| CN205721986U (en) | A kind of cable visual defects detection and character recognition device | |
| CN115452846A (en) | An online defect monitoring system for medical non-woven fabrics based on machine vision | |
| CN203365339U (en) | On-line detecting system for surface defect of cold rolling aluminum plate | |
| CN115100562A (en) | Intelligent monitoring system and method for equipment line based on video image and deep learning | |
| CN204027594U (en) | For gathering the device of bearing roller end face and chamfering image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20121010 |