WO2020103464A1 - 一种机房设备异常灯的识别方法以及识别系统 - Google Patents

一种机房设备异常灯的识别方法以及识别系统

Info

Publication number
WO2020103464A1
WO2020103464A1 PCT/CN2019/094510 CN2019094510W WO2020103464A1 WO 2020103464 A1 WO2020103464 A1 WO 2020103464A1 CN 2019094510 W CN2019094510 W CN 2019094510W WO 2020103464 A1 WO2020103464 A1 WO 2020103464A1
Authority
WO
WIPO (PCT)
Prior art keywords
object distance
camera
light
abnormal
preset
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.)
Ceased
Application number
PCT/CN2019/094510
Other languages
English (en)
French (fr)
Inventor
何东杰
杨洁
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Unionpay Co Ltd
Original Assignee
China Unionpay Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by China Unionpay Co Ltd filed Critical China Unionpay Co Ltd
Publication of WO2020103464A1 publication Critical patent/WO2020103464A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; 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
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; 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
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60—Control of cameras or camera modules
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60—Control of cameras or camera modules
    • H04N23/67—Focus control based on electronic image sensor signals
    • 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
    • Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
    • Y02B20/00—Energy efficient lighting technologies, e.g. halogen lamps or gas discharge lamps
    • Y02B20/40—Control techniques providing energy savings, e.g. smart controller or presence detection

Definitions

  • the present invention relates to image analysis technology, and in particular to a method and system for identifying abnormal lights in equipment rooms.
  • machine learning algorithms or deep learning algorithms are generally used to improve the effect of feature analysis.
  • the core of the existing technical solutions is to design better machine learning or deep learning algorithms to improve the extraction effect of signal lights and abnormal light features, thereby achieving a higher recognition rate.
  • the present invention aims to provide a method and a system for identifying abnormal lights of a computer room device that can optimize the photographing effect and reduce the influence of impurities on the photographing.
  • the object distance obtaining step focus on the cabinet door in the equipment room through the camera to obtain the corresponding object distance u;
  • the step of changing the object distance causes the camera to start changing based on the initial object distance and causes the camera to shoot with the changed object distance;
  • the judging step for the size of the light spot of the captured picture recognition signal lamp and the black proportion in the picture histogram, determine whether the light spot meets the preset first preset range and whether the black proportion in the picture histogram meets the second preset Specified range, if it is judged that the light spot does not meet the first preset specified range and the proportion of black in the slice histogram does not meet the second preset specified range, the step of changing the object distance is repeated until Comply with the first preset prescribed range and the second preset prescribed range; and
  • Recognition step analyze the picture and identify abnormal lights.
  • the camera takes u as the initial object distance and reduces the object distance by ⁇ , so that the camera shoots the cabinet door at the reduced object distance, where ⁇ is much smaller than u.
  • the camera is made to have an initial object distance of 0 and the object distance is increased by ⁇ , and the camera is caused to shoot at the increased object distance, where ⁇ is much smaller than u.
  • the first preset specified range refers to the size of the small hole of the cabinet door.
  • the picture is subjected to brightness detection to identify the signal lamp, and chromaticity analysis is performed to identify the signal lamp color, thereby determining the abnormal lamp.
  • the ⁇ is equal to 0.01u, and the second preset specified range is more than 90%.
  • a system for identifying abnormal lights in an equipment room includes:
  • the object distance obtaining module focuses the cabinet door in the equipment room through the camera to obtain the corresponding object distance u;
  • the object distance changing module causes the camera to start changing based on the initial object distance and causes the camera to shoot at the changed object distance;
  • the judgment module determines whether the spot size of the image recognition signal lamp obtained by shooting and the black proportion in the picture histogram meet the preset first preset range and whether the black proportion in the picture histogram meets the second preset Specified range, if it is determined that the light spot does not meet the first preset specified range and the black ratio in the slice histogram does not meet the second preset specified range, repeating the step of changing the object distance Act until the first preset specified range and the second preset specified range are met; and the identification module analyzes the picture taken by the camera and identifies the abnormal light.
  • the camera takes u as the initial object distance and reduces the object distance by ⁇ , so that the camera shoots the cabinet door at the reduced object distance, where ⁇ is much smaller than u.
  • the camera is made to have an initial object distance of 0 and the object distance is increased by ⁇ , and the camera is caused to shoot at the increased object distance, where ⁇ is much smaller than u.
  • the first preset specified range refers to the size of the small hole of the cabinet door.
  • the identification module performs brightness detection on the picture to identify the signal light and performs chromaticity analysis to identify the color of the signal light, thereby determining an abnormal light.
  • the ⁇ is equal to 0.01u, and the second preset specified range is more than 90%.
  • the computer-readable storage medium of the present invention has a computer program stored thereon, which is characterized in that when the program is executed by a processor, the above method for identifying abnormal lights of a computer room device is realized.
  • the computer device of the present invention includes a memory, a processor, and a computer program stored on the memory and capable of running on the processor, and is characterized in that, when the processor executes the program, the above-mentioned method for identifying abnormal lights of equipment rooms .
  • the photographing effect can be optimized, so that the abnormal light and the normal light are more prominent in the picture, and the influence of noise or impurities on the photograph recognition is greatly reduced, so that it can be more simple, efficient, and accurate It is recognized that there is no need for complex machine learning as in the prior art.
  • FIG. 1 is a flowchart showing a method for identifying an abnormal light in an equipment room according to an embodiment of the present invention.
  • FIG. 2 is a schematic diagram showing the structure of a system for identifying abnormal lights in an equipment room according to an embodiment of the present invention.
  • the machine room inspection robot In the identification of abnormal lights in the equipment room, generally, when the cabinet is opened, the machine room inspection robot directly takes pictures of the device. Due to the various colors on the device panel, there are a large number of areas that are relatively close to the color of the lamp , Making identification difficult. When the cabinet is closed, the machine room inspection robot takes a picture of the closed device. Because the focus is on the cabinet door, the cabinet door reflects light and the brightness is high, so it is impossible to effectively distinguish the lights of different colors. Therefore, the inventor found through the above research that when taking pictures of the device, it is necessary to adjust the focus of light diffraction and the aperture of the cabinet door to diffract the light to reduce the influence of the cabinet door reflection and the like.
  • the cabinet When the cabinet is opened or closed, take pictures of the equipment in the equipment room. Due to the weak light, the pictures have more noise, and the lighting is not obvious.
  • the object distance when the camera shoots is u.
  • the robot is stationary, and the object distance is reduced to continue taking pictures.
  • the device As the object distance is reduced, the device is out of the focus point, and the lights will be blurred, making it easier to identify.
  • the other parts of the device will become more blurred due to blur, and the whole will become gray and dark. Then, keep reducing the object distance, so that the black part in the photo histogram occupies most of the size.
  • the size of the blurred light is basically the same as the size of the small hole on the cabinet door.
  • the blurred light in the final picture will not affect the formation of light spots.
  • most of the other pictures are basically black or gray. You can use this picture to identify signal lights and abnormal lights.
  • the brightness and chromaticity analysis can easily and efficiently complete the identification of signal lights and abnormal lights. In this way, it is possible to prevent noise from affecting the extraction and recognition of signal lights and abnormal light features, and no complicated learning is required.
  • FIG. 1 is a flowchart showing a method for identifying an abnormal light in an equipment room according to an embodiment of the present invention.
  • a method for identifying abnormal lights in a computer room includes the following steps:
  • Step S1 When the cabinet is closed or opened, the machine room inspection robot focuses on the cabinet door through the camera to obtain the corresponding object distance u;
  • Step S2 Taking the object distance u as the initial object distance, reduce the object distance by ⁇ (where ⁇ is much smaller than u, for example, as an example, it can be assumed to be 0.01u), that is, reduce the object distance to u- ⁇ , and Shooting, in addition, as a modification, the object distance u may be used as the initial object distance, and the object distance may be reduced (u- ⁇ ), so that the camera shoots the cabinet door with the reduced object distance , Where ⁇ is much smaller than u;
  • Step S3 Analyze the captured picture, analyze the spot size of the identification signal lamp and the percentage of black in the picture histogram;
  • Step S4 If the light spot is too small to be effectively recognized and the black in the picture histogram is relatively small, then step S5 will be entered. In step S5, the object distance is reduced by ⁇ and shooting is performed, and then return to step S3, another
  • the size of the light spot is basically the same as the size of the small hole of the cabinet door and the black account is relatively high (for example, the black accounted for more than 90%), it indicates that the picture at the object distance can be used to identify the signal light and abnormal light;
  • Step S6 Perform brightness detection on the picture to identify the signal light, and perform chroma analysis to identify the color of the signal light to determine the abnormal light.
  • the method for recognizing abnormal lights in equipment rooms of this modification includes the following steps:
  • the size of the light spot is basically the same as the size of the small hole of the cabinet door and the proportion of black is relatively high (for example, more than 85%, or more than 90%, etc.), use the picture at the object distance to identify the signal light and abnormal light;
  • FIG. 2 is a schematic diagram showing the structure of a system for identifying abnormal lights in an equipment room according to an embodiment of the present invention.
  • a system for identifying abnormal lights in a computer room includes:
  • Camera 100 used to photograph cabinet doors
  • the object distance obtaining module 200 is used to obtain the object distance u for the camera to focus on the cabinet door in the equipment room;
  • the object distance reduction module 300 causes the camera to start changing based on the initial object distance and causes the camera to shoot at the changed object distance;
  • the judging module 400 determines whether the light spot size of the captured picture identification signal lamp and the black proportion in the picture histogram are within the first preset range and whether the black proportion in the picture histogram meets the second preset Set a prescribed range, and if it is determined that the light spot does not meet the first preset prescribed range and the proportion of black in the slice histogram does not meet the second preset prescribed range, repeat the object distance changing module The action of 300 until it meets the first preset specified range and the second preset specified range; and
  • the identification module 500 performs brightness detection on the picture to identify the signal light and performs chromaticity analysis to identify the color of the signal light, thereby determining the abnormal light.
  • the object distance changing module 300 changes the object distance Continue to decrease and allow the camera to shoot the cabinet door at the reduced object distance.
  • the object distance changing module 300 makes the camera take the initial object distance of 0 and increases the object distance by ⁇ , and causes the camera to shoot at the increased object distance, where ⁇ is much smaller than u.
  • the first preset specified range may be set to be substantially the same as the size of the small hole of the cabinet door, for example.
  • the second preset predetermined range may be set to a value of, for example, 85% or more, 90% or more, and so on.
  • the present invention also provides a computer-readable storage medium on which a computer program is stored, which is characterized in that, when the program is executed by a processor, the above method for identifying abnormal lights of a computer room equipment is realized.
  • the present invention also provides a computer device, including a memory, a processor, and a computer program stored on the memory and capable of running on the processor, characterized in that, when the processor executes the program, the foregoing abnormal light of the equipment room equipment is realized Identification method.
  • the machine room inspection robot directly takes pictures of the device. Since there are various colors on the device panel, there are a large number of areas that are relatively close to the color of the lamp, which makes the recognition difficult.
  • the machine room inspection robot takes a picture of the closed device. Because the focus is on the cabinet door, the cabinet door reflects light and the brightness is high, so it is impossible to effectively distinguish the lights of different colors. Therefore, the inventor found through the above research that when taking pictures of the device, it is necessary to adjust the focus of the light and the diffraction of the light on the cabinet door to reduce the influence of the cabinet door reflection and the like.
  • the method for identifying abnormal lights in a computer room and the system for identifying abnormal lights in a computer room of the present invention when taking pictures of the equipment in the cabinet when the cabinet is opened or closed, will cause the equipment to be out of the focus point due to the reduced object distance. It will be blurred and more easily recognized. Other parts of the device will become more blurred due to the blur, and the whole will become gray and dark. Next, keep reducing the object distance so that the black part of the photo histogram occupies the most part.
  • the size of the blurred light is basically the same as the size of the small hole on the cabinet door. There is no influence between the light spots formed by the light, and most of the other pictures are basically black or gray. Therefore, through simple brightness and chromaticity analysis of the picture, the identification of signal lights and abnormal lights can be completed simply and efficiently.
  • the photographing effect can be optimized, so that the abnormal light and the normal light are more prominent in the picture, and the influence of noise or impurities on the photograph recognition is greatly reduced, so that it can be more simple, efficient, and accurate It is recognized that there is no need for complex machine learning as in the prior art.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Signal Processing (AREA)
  • Image Analysis (AREA)
  • Studio Devices (AREA)

Abstract

本发明涉及机房设备异常灯的识别方法及其系统。该方法包括：物距改变步骤，使得摄像头的以初始物距为基础开始改变并且使得摄像机以该改变后的物距进行拍摄；判断步骤，对于拍摄得到的图片识别信号灯的光斑大小及图片直方图中的黑色占比，判断光斑是否符合预设第一预设规定范围以及图片直方图中的黑色占比是否符合第二预设规定范围，若判断所述光斑不符合第一预设规定范围以及图片直方图中的黑色占比不符合第二预设规定范围的情况下，重复进行物距改变步骤直到符合第一预设规定范围以及第二预设规定范围；以及识别步骤，对该图片进行分析并识别异常灯。根据本发明，能够优化拍照效果，使得异常灯和正常灯在图片中更加突出。

Description

一种机房设备异常灯的识别方法以及识别系统 技术领域
本发明涉及图像分析技术，特别涉及用于机房设备异常灯的识别方法以及识别系统。
背景技术
随着机房设备规模的不断增加，通过人工巡检来发现设备异常的方式存在较大的工作量。业界开始使用轮式机器人在机房中自动行走，通过摄像头对设备进行拍照，从而识别异常灯。
现有的巡检机器人一般通过摄像头的自动对焦方式完成对环境的拍照，随后再对图像进行深度分析和异常灯识别。典型的步骤如下：
通过摄像头对机房设备进行对焦拍照；
对拍摄的照片进行特征分析，一般采用机器学习算法或者深度学习算法，提高特征分析的效果。
由于机房环境光线、设备面板、机柜透光等原因，使得图片较暗、杂色较多以及信号灯不够明显。在该种情况下，现有技术方案中的核心在于设计较好的机器学习或者深度学习算法，提高信号灯和异常灯特征的提取效果，从而达到的较高的识别率。
然而，在这样的现有技术中，光线较弱会带来较多的杂色影响信号灯和异常灯特征的提取和识别，光线较强又会使得信号灯和异常灯的光线不够突出，影响灯的识别，显然光线的强弱对于灯光的识别影响较大，对机器学习或者深度学习算法的要求较高，难度较大。
公开于本发明背景部分的信息仅仅旨在增加对本发明的总体背景的理解，而不应当被视为承认或以任何形式暗示该信息构成已为本领域一般技术人员所公知的现有技术。
发明内容
鉴于上述问题，本发明旨在提供一种能够优拍照效果并减少杂质对拍照影响的机房设备异常灯的识别方法以及识别系统。
本发明的一方面的机房设备异常灯的识别方法，其特征在于，包括：
物距获得步骤，通过摄像头对机房设备中的机柜门进行对焦，获得相应的物距u；
物距改变步骤，使得摄像头的以初始物距为基础开始改变并且使得摄像机以该改变后的物距进行拍摄；
判断步骤，对于拍摄得到的图片识别信号灯的光斑大小以及图片直方图中的黑色占比，判断光斑是否符合预设第一预设规定范围以及图片直方图中的黑色占比是否符合第二预设规定范围，若判断所述光斑不符合所述第一预设规定范围以及所述片直方图中的黑色占比不符合第二预设规定范围的情况下，重复进行所述物距改变步骤直到符合所述第一预设规定范围以及所述第二预设规定范围；以及
识别步骤，对该图片进行分析并识别异常灯。
可选地，在物距改变步骤中，使得摄像头以u为初始物距并将物距减小δ，使得所述摄像头以该减小后的物距拍摄机柜门，其中，δ远小于u。
可选地，在物距改变步骤中，使得摄像头以0为初始物距并将物距增大δ，并且使得摄像机以该增大后的物距进行拍摄，其中，δ远小于u。
可选地，在所述判断步骤中，所述第一预设规定范围是指机柜门小孔大小。
可选地，在所述识别步骤中，对图片进行亮度检测而识别出信号灯，并且进行色度分析而识别信号灯颜色，由此确定异常灯。
可选地，所述δ等于0.01u，所述第二预设规定范围为90％以上。
本发明的一方面的机房设备异常灯的识别系统，其特征在于，包括：
摄像头，用于拍摄机柜门；
物距获得模块，通过摄像头对机房设备中的机柜门进行对焦，获得相应的物距u；
物距改变模块，使得摄像头的以初始物距为基础开始改变并且使得摄像机以该改变后的物距进行拍摄；
判断模块，对于拍摄得到的图片识别信号灯的光斑大小以及图片直方图中的黑色占比，判断光斑是否符合预设第一预设规定范围以及图片直方图中的黑色占比是否符合第二预设规定范围，若判断所述光斑不符合所述第一预设规定范围以及所述片直方图中的黑色占比不符合第二预设规定范围的情况下，重复进行所述物距改变模块的动作直到符合所述第一预设规定范围以及所述第二预设规定范围；以及识别模块，对所述摄像头拍摄到的图片进行分析并识别异常灯。
可选地，在物距改变模块中，使得摄像头以u为初始物距并将物距减小δ，使得所述摄像头以该减小后的物距拍摄机柜门，其中，δ远小于u。
可选地，在物距改变模块中，使得摄像头以0为初始物距并将物距增大δ，并且使得摄像机以该增大后的物距进行拍摄，其中，δ远小于u。
可选地，所述第一预设规定范围是指机柜门小孔大小。
可选地，所述识别模块对图片进行亮度检测而识别出信号灯并且进行色度分析而识别信号灯颜色，由此确定异常灯。
可选地，所述δ等于0.01u，所述第二预设规定范围为90％以上。
本发明的计算机可读存储介质，其上存储有计算机程序，其特征在于，该程序被处理器执行时实现上述的机房设备异常灯的识别方法。
本发明的计算机设备，包括存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，其特征在于，所述处理器执行所述程序时实现上述的机房设备异常灯的识别方法。
如上所述，根据通过本发明，能够优化拍照效果，使得异常灯和正常灯在图片中更加突出，较大程度减少了杂色或者杂质对拍照 识别的影响，从而能够更加简单、高效、精准的被识别出来，不需要如现有技术一般进行复杂的机器学习。
通过纳入本文的附图以及随后与附图一起用于说明本发明的某些原理的具体实施方式，本发明的方法和装置所具有的其它特征和优点将更为具体地变得清楚或得以阐明。
附图说明
图1是表示本发明一实施方式的机房设备异常灯的识别方法的流程图。
图2是表示本发明一实施方式的机房设备异常灯的识别系统的构造示意图。
具体实施方式
下面介绍的是本发明的多个实施例中的一些，旨在提供对本发明的基本了解。并不旨在确认本发明的关键或决定性的要素或限定所要保护的范围。
在机房设备异常灯的识别中，一般情况下，是在机柜开门时候，通过机房巡检机器人直接对着设备进行拍照，由于设备面板上有各种颜色，存在大量和灯的颜色比较接近的区域，使得识别难度大大提高。机柜关门时候，通过机房巡检机器人对着关门的设备进行拍照，由于对焦在机柜门上，机柜门反光使得亮度较高，无法有效区分不同颜色的灯。因此，本发明人经过上述研究发现，在对设备进行拍照时候，需要结合光照以及机柜门上小孔对光的衍射进行对焦点进行调整，以降低机柜门反光等的影响。
在机柜开门时候或者关门时候，对机房设备进行拍照，由于光线较弱，使得拍照出来的图片杂色较多，同时灯光不够明显。假设摄像头拍摄时候的物距为u。机器人固定不动，缩小物距继续拍照，由于物距缩小将使得设备处于对焦点之外，灯光将被虚化，更加容易被识别。设备其他部分由于虚化，将变得更加模糊，整体将变灰变暗。接着，不断缩小物距，使得照片直方图中的黑色部分占据绝 大部分以及在关闭机柜门时候，虚化的灯光大小基本和机柜门上的小孔大小一致。最终形成的图片中虚化的灯光形成光斑之间不会产生影响，同时图片其他大部分基本是黑色或者灰色，能够使用该图片进行信号灯和异常灯的识别，由此，能够通过对图片进行简单的亮度和色度分析，可以简单高效的完成信号灯和异常灯的识别。由此能够避免杂色影响信号灯和异常灯特征的提取和识别，也不需要进行复杂的学习。
接着对于本发明一实施方式的机房设备异常灯的识别方法进行说明。
图1是表示本发明一实施方式的机房设备异常灯的识别方法的流程图。
如图1所示，本发明一实施方式的机房设备异常灯的识别方法包括下述步骤：
步骤S1：在机柜关门或者开门时候，机房巡检机器人通过摄像头对机柜门进行对焦，获得相应的物距u；
步骤S2：以该物距u为初始物距，将物距缩小δ(其中，δ远小于u，例如作为一个示例，可假设为0.01u)，也就是减小物距为u-δ，并进行拍摄，另外，作为一个变形例，也可以是以该物距u为初始物距，而将物距减小(u-δ)，使得所述摄像头以该减小后的物距拍摄机柜门，其中，δ远小于u；
步骤S3：对拍摄的图片进行分析，分析识别信号灯的光斑大小以及图片直方图中的黑色占比；
步骤S4：如果光斑较小无法有效识别以及图片直方图中的黑色占比较小的情况下，则将进入步骤S5，在步骤S5中物距缩小δ并进行拍摄后，并返回步骤S3，另一方面，当光斑大小和机柜门小孔大小基本一致以及黑色占比较高时候(例如黑色占比达到90％以上)，表明可使用该物距下的图片进行信号灯和异常灯的识别；
步骤S6：对图片进行亮度检测识别出信号灯，以及进行色度分析识别信号灯颜色，从而确定异常灯。
接着对于上述实施方式的一个变形例进行说明。
该变形例的机房设备异常灯的识别方法包括下述步骤：
(1)在机柜关门或者开门时候，机房巡检机器人通过摄像头对机柜门进行对焦，获得相应的物距为u；
(2)以0为摄像头初始物距，将物距从0开始增大，例如增大δ(其中，δ远小于u，例如可假设为0.01u)，并进行拍照；
(3)对图片进行分析，分析识别信号灯的光斑大小以及图片直方图中的黑色占比，此时光斑较大并且明显大于机柜门小孔，并且黑色占比较大，继续增大物距并使得物距小于u；
(4)当光斑大小和机柜门小孔大小基本一致以及黑色占比较高(例如85％以上，或者90％以上等)情况下，使用该物距下的图片进行信号灯和异常灯的识别；
(5)对图片进行亮度检测识别出信号灯，以及进行色度分析识别信号灯颜色，从而确定异常灯。
以上对于本发明的机房设备异常灯的识别方法进行了说明，接着对于本发明的机房设备异常灯的识别系统进行说明。
图2是表示本发明一实施方式的机房设备异常灯的识别系统的构造示意图。
如图2所示，本发明的一实施方式的机房设备异常灯的识别系统具备：
摄像头100，用于拍摄机柜门；
物距获得模块200，用于获取摄像头对机房设备中的机柜门进行对焦的物距u；
物距减小模块300，使得摄像头的以初始物距为基础开始改变并且使得摄像机以该改变后的物距进行拍摄；
判断模块400，对于拍摄得到的图片识别信号灯的光斑大小以及图片直方图中的黑色占比，判断光斑是否符合预设第一预设规定范围以及图片直方图中的黑色占比是否符合第二预设规定范围，若判断所述光斑不符合所述第一预设规定范围以及所述片直方图中的黑色占比不符合第二预设规定范围的情况下，重复进行所述物距改变模块300的动作直到符合所述第一预设规定范围以及所述第二 预设规定范围；以及
识别模块500，对图片进行亮度检测而识别出信号灯并且进行色度分析而识别信号灯颜色，由此确定异常灯。
其中，物距改变模块300使得摄像头以u为初始物距并将物距减小δ，以使得所述摄像头以该减小后的物距拍摄机柜门，其中，δ远小于u，其中，例如δ＝0.01u。
也可以是，物距改变模块300使得摄像头以u为初始物距并将物距减小到δ，以使得摄像头以该减小后的物距拍摄机柜门，其中，δ远小于u，例如，δ＝0.01u。这种情况下，当判断模块500判断判断光斑不符合预设第一预设规定范围以及图片直方图中的黑色占不符合第二预设规定范围的情况下，物距改变模块300将物距继续减小并使得摄像头以该减小后的物距拍摄机柜门。
或者还有一种情况可以是，物距改变模块300使得摄像头以0为初始物距并将物距增大δ，并且使得摄像机以该增大后的物距进行拍摄，其中，δ远小于u。
这里，所述第一预设规定范围例如可以设定为与机柜门小孔大小基本一致。所述第二预设规定范围可以设定为例如85％以上、90％以上等的数值。
本发明还提供一种计算机可读存储介质，其上存储有计算机程序，其特征在于，该程序被处理器执行时实现上述的机房设备异常灯的识别方法。
本发明还提供一种计算机设备，包括存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，其特征在于，所述处理器执行所述程序时实现上述的机房设备异常灯的识别方法。
通过机房巡检机器人直接对着设备进行拍照，由于设备面板上有各种颜色，存在大量和灯的颜色比较接近的区域，使得识别难度大大提高。机柜关门时候，通过机房巡检机器人对着关门的设备进行拍照，由于对焦在机柜门上，机柜门反光使得亮度较高，无法有效区分不同颜色的灯。因此，本发明人经过上述研究发现，在对设备进行拍照时候，需要结合光照以及机柜门上小孔对光的衍射进行 对焦点进行调整，以降低机柜门反光等的影响。
本发明的机房设备异常灯的识别方法以及机房设备异常灯的识别系统，在机柜开门时候或者关门时候，对机房设备进行拍照的情况下，由于物距缩小将使得设备处于对焦点之外，灯光将被虚化，更加容易被识别，设备其他部分由于虚化，将变得更加模糊，整体将变灰变暗。接着，不断缩小物距，使得照片直方图中的黑色部分占据绝大部分以及在关闭机柜门时候，虚化的灯光大小基本和机柜门上的小孔大小一致，由此，图片中虚化的灯光形成的光斑之间不会产生影响，同时图片其他大部分基本是黑色或者灰色，由此能够通过对图片进行简单的亮度和色度分析，可以简单高效的完成信号灯和异常灯的识别。
如上所述，根据通过本发明，能够优化拍照效果，使得异常灯和正常灯在图片中更加突出，较大程度减少了杂色或者杂质对拍照识别的影响，从而能够更加简单、高效、精准的被识别出来，不需要如现有技术一般进行复杂的机器学习。
以上例子主要说明了本发明的机房设备异常灯的识别方法以及机房设备异常灯的识别系统。尽管只对其中一些本发明的具体实施方式进行了描述，但是本领域普通技术人员应当了解，本发明可以在不偏离其主旨与范围内以许多其他的形式实施。因此，所展示的例子与实施方式被视为示意性的而非限制性的，在不脱离如所附各权利要求所定义的本发明精神及范围的情况下，本发明可能涵盖各种的修改与替换。

Claims (14)

  1. 一种机房设备异常灯的识别方法，其特征在于，包括：
    物距获得步骤，通过摄像头对机房设备中的机柜门进行对焦，获得相应的物距u；
    物距改变步骤，使得摄像头的以初始物距为基础开始改变并且使得摄像机以该改变后的物距进行拍摄；
    判断步骤，对于拍摄得到的图片识别信号灯的光斑大小以及图片直方图中的黑色占比，判断光斑是否符合预设第一预设规定范围以及图片直方图中的黑色占比是否符合第二预设规定范围，若判断所述光斑不符合所述第一预设规定范围以及所述片直方图中的黑色占比不符合第二预设规定范围的情况下，重复进行所述物距改变步骤直到符合所述第一预设规定范围以及所述第二预设规定范围；以及
    识别步骤，对该图片进行分析并识别异常灯。
  2. 如权利要求1所述的机房设备异常灯的识别方法，其特征在于，
    在物距改变步骤中，使得摄像头以u为初始物距并将物距减小δ，使得所述摄像头以该减小后的物距拍摄机柜门，其中，δ远小于u。
  3. 如权利要求1所述的机房设备异常灯的识别方法，其特征在于，
    在物距改变步骤中，使得摄像头以0为初始物距并将物距增大δ，并且使得摄像机以该增大后的物距进行拍摄，其中，δ远小于u。
  4. 如权利要求1～3任意一项所述的机房设备异常灯的识别方法，其特征在于，
    在所述判断步骤中，所述第一预设规定范围是指机柜门小孔大小。
  5. 如权利要求1～3任意一项所述的机房设备异常灯的识别方法， 其特征在于，
    在所述识别步骤中，对图片进行亮度检测而识别出信号灯，并且进行色度分析而识别信号灯颜色，由此确定异常灯。
  6. 如权利要求1～3任意一项所述的机房设备异常灯的识别方法，其特征在于，
    所述δ等于0.01u，所述第二预设规定范围为90％以上。
  7. 一种机房设备异常灯的识别系统，其特征在于，包括：
    摄像头，用于拍摄机柜门；
    物距获得模块，通过摄像头对机房设备中的机柜门进行对焦，获得相应的物距u；
    物距改变模块，使得摄像头的以初始物距为基础开始改变并且使得摄像机以该改变后的物距进行拍摄；
    判断模块，对于拍摄得到的图片识别信号灯的光斑大小以及图片直方图中的黑色占比，判断光斑是否符合预设第一预设规定范围以及图片直方图中的黑色占比是否符合第二预设规定范围，若判断所述光斑不符合所述第一预设规定范围以及所述片直方图中的黑色占比不符合第二预设规定范围的情况下，重复进行所述物距改变模块的动作直到符合所述第一预设规定范围以及所述第二预设规定范围；以及
    识别模块，对所述摄像头拍摄到的图片进行分析并识别异常灯。
  8. 如权利要求7所述的机房设备异常灯的识别系统，其特征在于，
    在物距改变模块中，使得摄像头以u为初始物距并将物距减小δ，使得所述摄像头以该减小后的物距拍摄机柜门，其中，δ远小于u。
  9. 如权利要求7所述的机房设备异常灯的识别系统，其特征在于，
    在物距改变模块中，使得摄像头以0为初始物距并将物距增大δ，并且使得摄像机以该增大后的物距进行拍摄，其中，δ远小于u。
  10. 如权利要求7～9任意一项所述的机房设备异常灯的识别系 统，其特征在于，
    所述第一预设规定范围是指机柜门小孔大小。
  11. 如权利要求7～9任意一项所述的机房设备异常灯的识别系统，其特征在于，
    所述识别模块对图片进行亮度检测而识别出信号灯并且进行色度分析而识别信号灯颜色，由此确定异常灯。
  12. 如权利要求7～9任意一项所述的机房设备异常灯的识别系统，其特征在于，
    所述δ等于0.01u，所述第二预设规定范围为90％以上。
  13. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，该程序被处理器执行时实现权利要求1～6任意一项所述的机房设备异常灯的识别方法。
  14. 一种计算机设备，包括存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，其特征在于，所述处理器执行所述程序时实现所述权利要求1～6任意一项所述的机房设备异常灯的识别方法。
PCT/CN2019/094510 2018-11-22 2019-07-03 一种机房设备异常灯的识别方法以及识别系统 Ceased WO2020103464A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811398266.0A CN111209782B (zh) 2018-11-22 2018-11-22 一种机房设备异常灯的识别方法以及识别系统
CN201811398266.0 2018-11-22

Publications (1)

Publication Number Publication Date
WO2020103464A1 true WO2020103464A1 (zh) 2020-05-28

Family

ID=70774424

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/094510 Ceased WO2020103464A1 (zh) 2018-11-22 2019-07-03 一种机房设备异常灯的识别方法以及识别系统

Country Status (2)

Country Link
CN (1) CN111209782B (zh)
WO (1) WO2020103464A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112215106A (zh) * 2020-09-29 2021-01-12 国网上海市电力公司 一种用于变电站无人巡检系统的仪表颜色状态识别方法
CN112364740A (zh) * 2020-10-30 2021-02-12 交控科技股份有限公司 一种基于计算机视觉的无人机房监控方法及系统
CN116342459A (zh) * 2021-12-17 2023-06-27 沈阳新松机器人自动化股份有限公司 一种针对低分辨率的机房服务器故障灯检测方法

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115311283B (zh) * 2022-10-12 2023-01-24 山东鲁玻玻璃科技有限公司 一种玻璃拉管缺陷检测方法及系统
CN119095241A (zh) * 2024-07-18 2024-12-06 广州云能科技有限公司 灯光设备的故障检测方法及灯光控制系统

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004241052A (ja) * 2003-02-06 2004-08-26 Canon Inc 光磁気再生方法
CN103674839A (zh) * 2013-11-12 2014-03-26 清华大学 一种基于光斑检测的可视化样品定位操作系统及方法
CN105100732A (zh) * 2015-08-26 2015-11-25 深圳市银之杰科技股份有限公司 一种机房服务器远程监控方法及系统
CN108241366A (zh) * 2016-12-27 2018-07-03 中国移动通信有限公司研究院 一种移动巡检机器人和移动巡检系统

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105578026B (zh) * 2015-07-10 2017-11-17 宇龙计算机通信科技(深圳)有限公司 一种拍摄方法及用户终端
CN105611140A (zh) * 2015-07-31 2016-05-25 宇龙计算机通信科技(深圳)有限公司 拍照控制方法、拍照控制装置及终端
CN105163042B (zh) * 2015-08-03 2017-11-03 努比亚技术有限公司 一种虚化处理深度图像的装置和方法
CN106550184B (zh) * 2015-09-18 2020-04-03 中兴通讯股份有限公司 照片处理方法及装置
CN106161980A (zh) * 2016-07-29 2016-11-23 宇龙计算机通信科技(深圳)有限公司 基于双摄像头的拍照方法及系统
CN107026979B (zh) * 2017-04-19 2020-09-11 宇龙计算机通信科技(深圳)有限公司 双摄像头拍照方法及装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004241052A (ja) * 2003-02-06 2004-08-26 Canon Inc 光磁気再生方法
CN103674839A (zh) * 2013-11-12 2014-03-26 清华大学 一种基于光斑检测的可视化样品定位操作系统及方法
CN105100732A (zh) * 2015-08-26 2015-11-25 深圳市银之杰科技股份有限公司 一种机房服务器远程监控方法及系统
CN108241366A (zh) * 2016-12-27 2018-07-03 中国移动通信有限公司研究院 一种移动巡检机器人和移动巡检系统

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112215106A (zh) * 2020-09-29 2021-01-12 国网上海市电力公司 一种用于变电站无人巡检系统的仪表颜色状态识别方法
CN112364740A (zh) * 2020-10-30 2021-02-12 交控科技股份有限公司 一种基于计算机视觉的无人机房监控方法及系统
CN112364740B (zh) * 2020-10-30 2024-04-19 交控科技股份有限公司 一种基于计算机视觉的无人机房监控方法及系统
CN116342459A (zh) * 2021-12-17 2023-06-27 沈阳新松机器人自动化股份有限公司 一种针对低分辨率的机房服务器故障灯检测方法

Also Published As

Publication number Publication date
CN111209782B (zh) 2024-04-16
CN111209782A (zh) 2020-05-29

Similar Documents

Publication Publication Date Title
WO2020103464A1 (zh) 一种机房设备异常灯的识别方法以及识别系统
CN103617432B (zh) 一种场景识别方法及装置
EP3579145A1 (en) Method and device for image processing, computer readable storage medium, and electronic device
TWI419082B (zh) 偵測視訊序列中的移動物體影像的方法以及影像處理系統
AU2010238543B2 (en) Method for video object detection
CN116320771B (zh) 一种拍摄方法和电子设备
AU2011203219B2 (en) Mode removal for improved multi-modal background subtraction
CN108734676A (zh) 图像处理方法和装置、电子设备、计算机可读存储介质
CN103679704B (zh) 一种基于光照补偿的视频运动阴影检测方法
CN112598746B (zh) 电梯门开关检测方法、装置、可读介质及电子设备
CN103905727A (zh) 被摄体区域跟踪设备及其控制方法
US20160286114A1 (en) Image capture by scene classification
CN110830794A (zh) 光源检测方法、终端及存储介质
US20110129149A1 (en) Method and apparatus for processing digital image signal, recording medium having recorded thereon the method, and apparatus for processing digital image signal
CN101364304A (zh) 基于颜色不变性和高斯模型的阴影检测方法
CN101213828A (zh) 将虹膜颜色结合到红眼修正中的方法和设备
US11195298B2 (en) Information processing apparatus, system, method for controlling information processing apparatus, and non-transitory computer readable storage medium
CN111160299A (zh) 活体识别方法及装置
CN113808135B (zh) 图像亮度异常检测方法、电子设备及存储介质
WO2023045836A1 (zh) 灯具检测方法、装置、设备、介质、芯片、产品及程序
CN112770021B (zh) 摄影机与滤片切换方法
CN109995984A (zh) 一种照明方法、装置、电子设备及存储介质
CN105898156A (zh) 一种基于降低亮度值的拍照方法和系统
CN108629329A (zh) 图像处理方法和装置、电子设备、计算机可读存储介质
CN111147693B (zh) 一种全尺寸拍照图像的降噪方法及装置

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: 19887538

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: 19887538

Country of ref document: EP

Kind code of ref document: A1