WO2017080283A1 - 一种图像校准方法及装置 - Google Patents
一种图像校准方法及装置 Download PDFInfo
- Publication number
- WO2017080283A1 WO2017080283A1 PCT/CN2016/096893 CN2016096893W WO2017080283A1 WO 2017080283 A1 WO2017080283 A1 WO 2017080283A1 CN 2016096893 W CN2016096893 W CN 2016096893W WO 2017080283 A1 WO2017080283 A1 WO 2017080283A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- circuit board
- calibration
- feature
- point
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/001—Industrial image inspection using an image reference approach
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30141—Printed circuit board [PCB]
Definitions
- the present invention relates to the field of automatic optical detection technologies, and in particular, to an image calibration method and apparatus.
- Automated Optical Inspection is an effective method for industrial automation. It uses machine vision as a standard for inspection. It is widely used in LCD/TFT, transistor and PCB industry processes, and can be extended to the security system for people's death. Automated optical inspection is a common method commonly used in industrial processes. It uses optical methods to obtain the surface state of finished products, and image processing to detect foreign matter or pattern anomalies. Because it is a non-contact inspection, semi-finished products can be inspected in intermediate works.
- the method of calibrating an image is generally to design a special positioning point on the circuit board. During the image calibration process, the positioning point is collected, and the circuit board image is calibrated according to the displacement of the positioning point.
- the above method will cause the production point and recognition error of the anchor point to affect the calibration effect of the board image, and the acquisition of the anchor point is also susceptible to the production process, lossless and acquisition environment, and if the board is not designed during positioning. At the point, it is impossible to calibrate and verify the board without the positioning point.
- the design of the positioning point will be affected by the designer's design experience, thus affecting the calibration accuracy of the board image.
- the embodiment of the invention provides an image calibration method and device, which can improve the calibration precision of the circuit board image.
- An embodiment of the present invention provides an image calibration method, including:
- the sign is the picture area in the standard image;
- the method further includes:
- the feature information includes picture and position information of the feature point; N ⁇ 2;
- a matching rule is configured according to the number of feature points; the matching rule includes feature information of the feature point, an image conversion type, and an image conversion degree.
- the image of the circuit board to be calibrated is adjusted according to a preset matching rule, so that the positioning point is the same as the position of the feature point, and specifically includes:
- the image of the circuit board to be calibrated is adjusted within a range of the image conversion degree so that each positioning point has the same position as its corresponding feature point.
- the method further includes:
- the image conversion type includes at least one of a rotation, an affine transformation, or a perspective transformation; the image conversion degree includes an image angle conversion degree and an image scale conversion degree.
- an image calibration apparatus including:
- An image acquisition module configured to acquire an image of the circuit board to be calibrated
- a positioning point identification module configured to identify an positioning point from the image of the circuit board to be calibrated according to the pre-stored feature point; the feature point is a picture area in the standard image;
- the image adjustment module is configured to adjust an image of the circuit board to be calibrated according to a preset matching rule, so that the positioning point is the same as the position of the feature point.
- the image calibration apparatus further includes:
- a standard image acquisition module for acquiring a standard image of the circuit board
- a feature point selection module configured to select N feature points from the standard image, and save feature information of each feature point correspondingly; the feature information includes a picture and location information of the feature point; N ⁇ 2;
- the rule configuration module is configured to configure a matching rule according to the number of feature points; the matching rule includes feature information of the feature point, an image conversion type, and an image conversion degree.
- the image adjustment module is specifically configured to adjust an image of the circuit board to be calibrated within a range of the image conversion degree according to the image conversion type, so that each positioning point has the same position as its corresponding feature point.
- the image calibration apparatus further includes:
- a calibration test module configured to perform calibration test on the sampled image of the circuit board by using the matching rule
- a rule adjustment module configured to adjust the number of feature points and corresponding matching rules if the calibration test fails, and continue the calibration test;
- the saving module is configured to save the matching rule if the calibration test passes.
- the image conversion type includes at least one of a rotation, an affine transformation, or a perspective transformation; the image conversion degree includes an image angle conversion degree and an image scale conversion degree.
- the image calibration method and device provided by the embodiments of the present invention can take a picture area with unique features in a standard image as a feature point, and identify an anchor point from the image of the circuit board to be calibrated according to the feature point, according to the positioning point and the feature.
- the positional relationship of the points, the matching rule is used to adjust the image of the circuit board to be calibrated, so that the positions of the components in the image of the circuit board to be calibrated are consistent with those in the standard image, thereby improving the calibration accuracy of the circuit board image without the need for circuit board design.
- Set the positioning point to improve the applicability of image calibration for the non-locating board.
- different matching rules are configured, so that different image conversion types are used to adjust the image of the circuit board to be calibrated, and the applicability and robustness of the device are improved; after the matching rules are configured, the method is adopted.
- the matching rule performs a calibration test on the sampled pattern. Matching rules can be adjusted to reduce calibration errors.
- FIG. 1 is a schematic flow chart of an embodiment of an image calibration method provided by the present invention.
- FIG. 2 is a partial flow chart of another embodiment of an image calibration method provided by the present invention.
- FIG. 3 is a schematic structural view of an embodiment of an image calibration apparatus provided by the present invention.
- FIG. 4 is a partial structural schematic view of another embodiment of an image calibration apparatus provided by the present invention.
- FIG. 1 is a schematic flowchart of an embodiment of an image calibration method provided by the present invention, including:
- the calibration circuit board is first photographed, and then the corresponding positioning point is identified from the image of the circuit board to be calibrated according to the pre-stored picture of the feature point.
- the feature points are picture areas in the standard image that are less repeated, are not easily confused with other positions, and have unique features.
- the image of the circuit board to be calibrated is adjusted so that the position of the positioning point and its corresponding feature point are the same, thereby aligning the image calibration of the circuit board to be calibrated with the standard image, thereby improving the calibration accuracy of the circuit board image. .
- the method before the acquiring the image of the circuit board to be calibrated, the method further includes:
- S02 selecting N feature points from the standard image, and correspondingly storing feature information of each feature point; the feature information includes picture and location information of the feature point; N ⁇ 2;
- the matching rule includes feature information of the feature point, an image conversion type, and an image conversion degree.
- the matching rules of the board image must be configured. Take a standard image of the board and take one or more photos. N feature points are selected from the standard image. The selection criterion of the feature points is a picture area with unique features in the standard image, and multiple feature points can be selected at one time as spare feature points. After selecting feature points, different matching rules can be configured according to the number of feature points. For example, if the number of feature points is 2, the image conversion type is set to rotation, that is, the image of the circuit board to be calibrated can be rotated during the calibration process; if the number of feature points is 3, the image is set.
- the conversion types are affine transformation and perspective transformation, that is, the affine transformation and perspective transformation of the image to be calibrated during the calibration process.
- affine transformation and perspective transformation that is, the affine transformation and perspective transformation of the image to be calibrated during the calibration process.
- the feature points are sorted and their corresponding feature information is saved to the matching rule.
- multiple matching rules can be configured for the same standard image, and the priority of each matching rule is set. In the calibration process, the matching rule is selected according to the priority to calibrate the image.
- the image of the circuit board to be calibrated is adjusted according to a preset matching rule, so that the positioning point is the same as the position of the feature point, and specifically includes:
- the image of the circuit board to be calibrated is adjusted within a range of the image conversion degree so that each positioning point has the same position as its corresponding feature point.
- the image to be calibrated is converted, and the conversion range is the range of the degree of image conversion set in the matching rule.
- the positioning points in the image of the circuit board to be calibrated are respectively corresponding to the feature points according to the sequence of the feature points in the matching rule, so that each positioning point is the same as the position of the corresponding feature point after the conversion.
- the method further includes:
- the matching rule can be used to perform calibration test on the sampling image of the circuit board, and the calibration accuracy is tested, and the test score is given.
- the test method first identifies the positioning point from the sampled image according to the feature point, and then performs calibration according to the positioning point.
- the test score includes the number of anchor points, the accuracy of the anchor point recognition, and the distance of the anchor point. Among them, the more the number of positioning points, the higher the image calibration accuracy, the higher the test score; after adding noise in the sampled image, or reducing the quality of the sampled image, the accuracy of the identification of the positioning point is scored; the larger the distance of the positioning point, the positioning The more accurate the point recognition, the higher the test score.
- the score After the score is over, judge whether the calibration test passes or not according to the test score. If the calibration test passes, the result of the test using the matching rule is accurate.
- the matching rule can be used to test the image of the calibration board; if the calibration test is not Passing, it indicates that the result of the test using the matching rule is inaccurate, and the selected feature points can be adjusted, and the image conversion type and image conversion degree in the matching rule can be used to achieve satisfactory results by using the matching rule for calibration.
- the image conversion type includes at least one of a rotation, an affine transformation, or a perspective transformation; the image conversion degree includes an image angle conversion degree and an image scale conversion degree.
- the image calibration method provided by the embodiment of the invention can take a picture area with unique features in the standard image as a feature point, and identify an anchor point from the image of the circuit board to be calibrated according to the feature point, according to the positioning point and the feature point. Positional relationship, using matching rules to adjust the image of the board to be calibrated, so that the position of each component in the image of the board to be calibrated is consistent with that in the standard image, improving the calibration accuracy of the board image, and eliminating the need to set the board design
- the positioning point improves the applicability of the image calibration of the board without the positioning point; according to the number of different feature points, different matching rules are configured, so that different image conversion types are used to adjust the image of the circuit board to be calibrated, and the device is improved. Applicability and robustness; after the matching rules are configured, the matching rules are used to perform calibration tests on the sampling patterns, and the matching rules that cannot pass the calibration can be adjusted to reduce the calibration error.
- the present invention also provides an image calibration apparatus capable of implementing all the processes of the image calibration method in the above embodiment.
- FIG. 3 is a schematic structural diagram of an embodiment of an image calibration apparatus provided by the present invention, including:
- An image acquisition module 1 configured to acquire an image of a circuit board to be calibrated
- the positioning point identification module 2 is configured to identify an positioning point from the image of the circuit board to be calibrated according to the pre-stored feature point; the feature point is a picture area in the standard image;
- the image adjustment module 3 is configured to adjust an image of the circuit board to be calibrated according to a preset matching rule, so that the positioning point is the same as the position of the feature point.
- the image calibration apparatus further includes:
- a standard image acquisition module 01 configured to acquire a standard image of the circuit board
- the feature point selection module 02 is configured to select N feature points from the standard image, and save feature information of each feature point correspondingly; the feature information includes picture and position information of the feature point; N ⁇ 2;
- the rule configuration module 03 is configured to configure a matching rule according to the number of feature points; the matching rule includes feature information of the feature point, an image conversion type, and an image conversion degree.
- the image adjustment module is specifically configured to adjust an image of the circuit board to be calibrated within a range of the image conversion degree according to the image conversion type, so that each positioning point has the same position as its corresponding feature point.
- the image calibration apparatus further includes:
- a calibration test module configured to perform calibration test on the sampled image of the circuit board by using the matching rule
- a rule adjustment module configured to adjust the number of feature points and corresponding matching rules if the calibration test fails, and continue the calibration test;
- the saving module is configured to save the matching rule if the calibration test passes.
- the image conversion type includes at least one of a rotation, an affine transformation, or a perspective transformation; the image conversion degree includes an image angle conversion degree and an image scale conversion degree.
- the image calibration apparatus can take a picture area with unique features in the standard image as a feature point, and identify an positioning point from the image of the circuit board to be calibrated according to the feature point, according to the positioning point and the feature point. Positional relationship, using matching rules to adjust the image of the board to be calibrated, so that the position of each component in the image of the board to be calibrated is consistent with that in the standard image, improving the calibration accuracy of the board image, and eliminating the need to set the board design
- the positioning point improves the applicability of the image calibration of the board without the positioning point; according to the number of different feature points, different matching rules are configured, so that different image conversion types are used to adjust the image of the circuit board to be calibrated, and the device is improved. Applicability and robustness; after the matching rules are configured, the matching rules are used to perform calibration tests on the sampling patterns, and the matching rules that cannot pass the calibration can be adjusted to reduce the calibration error.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
本发明公开了一种图像校准方法，包括：获取待校准电路板的图像；根据预存的特征点，从所述待校准电路板的图像中识别出定位点；所述特征点为标准图像中的图片区域；根据预设的匹配规则，调整所述待校准电路板的图像，使所述定位点与所述特征点的位置相同。相应的，本发明还公开了一种图像校准装置。采用本发明实施例，能够提高电路板图像的校准精度。
Description
本发明涉及自动光学检测技术领域，尤其涉及一种图像校准方法及装置。
自动光学检测(AOI，Automated Optical Inspection)为工业自动化有效的检测方法，使用机器视觉作为检测标准技术，大量应用于LCD/TFT、晶体管与PCB工业制程上，在民生用途则可延伸至保全系统。自动光学检测是工业制程中常见的代表性手法，利用光学方式取得成品的表面状态，以影像处理来检出异物或图案异常等瑕疵，因为是非接触式检查，所以可在中间工程检查半成品。
在通过AOI获取电路板的图像，并对其图像进行检验前，需先将其图像校准为检测所需的标准图像。现有技术中，对图像进行校准的方法一般为在电路板上设计特殊定位点，在图像校准过程中，对定位点进行采集，并根据该定位点的位移来校准电路板图像。但是，上述方法会使定位点的生产和识别误差影响电路板图像的校准效果，而定位点的采集也容易受到生产过程、无损和采集环境的影响，而且，若电路板在设计时未加入定位点，则无法对没有定位点的电路板进行校准和检验，同时，定位点的设计会受到设计者的设计经验的影响，从而影响电路板图像的校准精度。
发明内容
本发明实施例提出一种图像校准方法及装置，能够提高电路板图像的校准精度。
本发明实施例提供一种图像校准方法，包括：
获取待校准电路板的图像；
根据预存的特征点，从所述待校准电路板的图像中识别出定位点；所述特
征点为标准图像中的图片区域；
根据预设的匹配规则，调整所述待校准电路板的图像，使所述定位点与所述特征点的位置相同。
进一步地，在所述获取待校准电路板的图像之前，还包括：
获取电路板的标准图像；
从所述标准图像中选取N个特征点，并相应保存每个特征点的特征信息；所述特征信息包括特征点的图片和位置信息；N≥2；
根据特征点的个数，配置匹配规则；所述匹配规则包括特征点的特征信息、图像转换类型和图像转换程度。
进一步地，所述根据预设的匹配规则，调整所述待校准电路板的图像，使所述定位点与所述特征点的位置相同，具体包括：
根据所述图像转换类型，在所述图像转换程度的范围内调整所述待校准电路板的图像，使每个定位点与其相应特征点的位置相同。
进一步地，在所述根据特征点的个数，配置匹配规则之后，还包括：
采用所述匹配规则，对电路板的采样图像进行校准测试；
若校准测试不通过，则调整特征点的个数及其相应的匹配规则，并继续进行校准测试；
若校准测试通过，则保存所述匹配规则。
进一步地，所述图像转换类型包括旋转、仿射变换或透视变换中的至少一个；所述图像转换程度包括图像角度转换程度和图像比例转换程度。
相应的，本发明还公开了一种图像校准装置，包括：
图像获取模块，用于获取待校准电路板的图像；
定位点识别模块，用于根据预存的特征点，从所述待校准电路板的图像中识别出定位点；所述特征点为标准图像中的图片区域；以及，
图像调整模块，用于根据预设的匹配规则，调整所述待校准电路板的图像，使所述定位点与所述特征点的位置相同。
进一步地，所述图像校准装置还包括：
标准图像获取模块，用于获取电路板的标准图像；
特征点选取模块，用于从所述标准图像中选取N个特征点，并相应保存每个特征点的特征信息；所述特征信息包括特征点的图片和位置信息；N≥2；以及，
规则配置模块，用于根据特征点的个数，配置匹配规则；所述匹配规则包括特征点的特征信息、图像转换类型和图像转换程度。
进一步地，所述图像调整模块具体用于根据所述图像转换类型，在所述图像转换程度的范围内调整所述待校准电路板的图像，使每个定位点与其相应特征点的位置相同。
进一步地，所述图像校准装置还包括：
校准测试模块，用于采用所述匹配规则，对电路板的采样图像进行校准测试；
规则调整模块，用于若校准测试不通过，则调整特征点的个数及其相应的匹配规则，并继续进行校准测试；以及，
保存模块，用于若校准测试通过，则保存所述匹配规则。
进一步地，所述图像转换类型包括旋转、仿射变换或透视变换中的至少一个；所述图像转换程度包括图像角度转换程度和图像比例转换程度。
实施本发明实施例，具有如下有益效果：
本发明实施例提供的图像校准方法及装置，能够将标准图像中特征具有唯一性的图片区域作为特征点，并根据特征点从待校准电路板的图像中识别出定位点，根据定位点和特征点的位置关系，采用匹配规则调整待校准电路板的图像，使待校准电路板的图像中各个元件的位置与标准图像中的相一致，提高电路板图像的校准精度，且无需在电路板设计时设置定位点，提高对无定位点电路板的图像校准的适用性。
而且，根据不同的特征点个数，配置不同的匹配规则，从而采用不同的图像转换类型来调整待校准电路板的图像，提高装置的适用性和鲁棒性；在配置好匹配规则后，采用匹配规则对采样图形进行校准测试，对于校准不通过的匹
配规则可进行调整，降低校准误差。
图1是本发明提供的图像校准方法的一个实施例的流程示意图；
图2是本发明提供的图像校准方法的另一个实施例的部分流程示意图；
图3是本发明提供的图像校准装置的一个实施例的结构示意图；
图4是本发明提供的图像校准装置的另一个实施例的部分结构示意图。
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
参见图1，是本发明提供的图像校准方法的一个实施例的流程示意图，包括：
S1、获取待校准电路板的图像；
S2、根据预存的特征点，从所述待校准电路板的图像中识别出定位点；所述特征点为标准图像中的图片区域；
S3、根据预设的匹配规则，调整所述待校准电路板的图像，使所述定位点与所述特征点的大小和位置相同。
需要说明的是，在校准过程中，先对待校准电路板进行拍照，再根据预存的特征点的图片，从拍摄待校准电路板的图像中识别出相应的定位点。其中，特征点为标准图像中重复较少、不容易与其他位置混淆、特征具有唯一性的图片区域。根据预设的匹配规则，对待校准电路板的图像进行调整，使定位点与其相应的特征点的位置相同，从而将待校准电路板的图像校准与标准图像对齐，提高了电路板图像的校准精度。
需要说明的是，在对电路板图像进行校准之后，电路板图像中各个元件的位置与标准图像中的相一致，方便后续对电路板图像中的各个待检测点进行识
别和检测。
进一步地，如图2所示，在所述获取待校准电路板的图像之前，还包括：
S01、获取电路板的标准图像；
S02、从所述标准图像中选取N个特征点，并相应保存每个特征点的特征信息；所述特征信息包括特征点的图片和位置信息；N≥2；
S03、根据特征点的个数，配置匹配规则；所述匹配规则包括特征点的特征信息、图像转换类型和图像转换程度。
需要说明的是，在对待校准电路板的图像进行校准前，需先配置电路板图像的匹配规则。先拍摄电路板的标准图像，可以拍摄一张或多张照片。从标准图像中选取N个特征点，特征点的选取准则为标准图像中特征具有唯一性的图片区域，一次可选取多个特征点以作为备用特征点。选取特征点后，可根据特征点的个数配置不同的匹配规则。例如，若特征点的个数为2，则设定图像转换类型为旋转，即在校准过程中，可对待校准电路板的图像进行旋转操作；若特征点的个数为3，则设定图像转换类型为仿射变换和透视变换，即在校准过程中，可对待校准电路板的图像进行仿射变换和透视变换的操作。通过对特征点的个数和图像转换类型的限定，可以避免错误的转换结果。同时，将特征点进行排序，并保存其相应特征信息到匹配规则中。而且，对于同一个标准图像可以配置多个匹配规则，并设置每个匹配规则的优先级，在校准过程中，根据优先级选取匹配规则对图像进行校准。
进一步地，所述根据预设的匹配规则，调整所述待校准电路板的图像，使所述定位点与所述特征点的位置相同，具体包括：
根据所述图像转换类型，在所述图像转换程度的范围内调整所述待校准电路板的图像，使每个定位点与其相应特征点的位置相同。
需要说明的是，根据匹配规则中设置的图像转换类型，对待校准电路板的图像进行转换，其转换范围为匹配规则中设置的图像转换程度的范围。转换过程中，按照匹配规则中特征点的序列，分别将待校准电路板的图像中的定位点与特征点相对应，使得转换后每个定位点均与其相应的特征点的位置相同。
进一步地，在所述根据特征点的个数，配置匹配规则之后，还包括：
采用所述匹配规则，对电路板的采样图像进行校准测试；
若校准测试不通过，则调整特征点的个数及其相应的匹配规则，并继续进行校准测试；
若校准测试通过，则保存所述匹配规则。
需要说明的是，在配置完匹配规则后，可采用该匹配规则对电路板的采样图像进行校准测试，测试其校准精度，并给出测试评分。其中，测试方法为先根据特征点从采样图像中识别出定位点，再根据定位点进行校准。测试评分包括定位点的个数、定位点识别精度、定位点距离等方面。其中，定位点数量越多，图像校准精度越高，则测试评分越高；在采样图像中增加噪声，或降低采样图像质量后，对定位点识别准确性进行评分；定位点距离越大，定位点识别越精准，则测试评分越高。评分结束后，根据测试评分来判断校准测试通过与否，若校准测试通过，则说明采用该匹配规则进行测试的结果精准，可使用该匹配规则对待校准电路板的图像进行测试；若校准测试不通过，则说明采用该匹配规则进行测试的结果不精准，可调节选取的特征点，以及匹配规则中的图像转换类型和图像转换程度，以使采用该匹配规则进行校准能达到满意的效果。
进一步地，所述图像转换类型包括旋转、仿射变换或透视变换中的至少一个；所述图像转换程度包括图像角度转换程度和图像比例转换程度。
本发明实施例提供的图像校准方法，能够将标准图像中特征具有唯一性的图片区域作为特征点，并根据特征点从待校准电路板的图像中识别出定位点，根据定位点和特征点的位置关系，采用匹配规则调整待校准电路板的图像，使待校准电路板的图像中各个元件的位置与标准图像中的相一致，提高电路板图像的校准精度，且无需在电路板设计时设置定位点，提高对无定位点电路板的图像校准的适用性；根据不同的特征点个数，配置不同的匹配规则，从而采用不同的图像转换类型来调整待校准电路板的图像，提高装置的适用性和鲁棒性；在配置好匹配规则后，采用匹配规则对采样图形进行校准测试，对于校准不通过的匹配规则可进行调整，降低校准误差。
相应的，本发明还提供一种图像校准装置，能够实现上述实施例中的图像校准方法的所有流程。
参见图3，是本发明提供的图像校准装置的一个实施例的结构示意图，包括：
图像获取模块1，用于获取待校准电路板的图像；
定位点识别模块2，用于根据预存的特征点，从所述待校准电路板的图像中识别出定位点；所述特征点为标准图像中的图片区域；以及，
图像调整模块3，用于根据预设的匹配规则，调整所述待校准电路板的图像，使所述定位点与所述特征点的位置相同。
进一步地，如图4所示，所述图像校准装置还包括：
标准图像获取模块01，用于获取电路板的标准图像；
特征点选取模块02，用于从所述标准图像中选取N个特征点，并相应保存每个特征点的特征信息；所述特征信息包括特征点的图片和位置信息；N≥2；以及，
规则配置模块03，用于根据特征点的个数，配置匹配规则；所述匹配规则包括特征点的特征信息、图像转换类型和图像转换程度。
进一步地，所述图像调整模块具体用于根据所述图像转换类型，在所述图像转换程度的范围内调整所述待校准电路板的图像，使每个定位点与其相应特征点的位置相同。
进一步地，所述图像校准装置还包括：
校准测试模块，用于采用所述匹配规则，对电路板的采样图像进行校准测试；
规则调整模块，用于若校准测试不通过，则调整特征点的个数及其相应的匹配规则，并继续进行校准测试；以及，
保存模块，用于若校准测试通过，则保存所述匹配规则。
进一步地，所述图像转换类型包括旋转、仿射变换或透视变换中的至少一个；所述图像转换程度包括图像角度转换程度和图像比例转换程度。
本发明实施例提供的图像校准装置，能够将标准图像中特征具有唯一性的图片区域作为特征点，并根据特征点从待校准电路板的图像中识别出定位点，根据定位点和特征点的位置关系，采用匹配规则调整待校准电路板的图像，使待校准电路板的图像中各个元件的位置与标准图像中的相一致，提高电路板图像的校准精度，且无需在电路板设计时设置定位点，提高对无定位点电路板的图像校准的适用性；根据不同的特征点个数，配置不同的匹配规则，从而采用不同的图像转换类型来调整待校准电路板的图像，提高装置的适用性和鲁棒性；在配置好匹配规则后，采用匹配规则对采样图形进行校准测试，对于校准不通过的匹配规则可进行调整，降低校准误差。
以上所述是本发明的优选实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本发明原理的前提下，还可以做出若干改进和润饰，这些改进和润饰也视为本发明的保护范围。
Claims (10)
- 一种图像校准方法，其特征在于，包括：获取待校准电路板的图像；根据预存的特征点，从所述待校准电路板的图像中识别出定位点；所述特征点为标准图像中的图片区域；根据预设的匹配规则，调整所述待校准电路板的图像，使所述定位点与所述特征点的位置相同。
- 如权利要求1所述的图像校准方法，其特征在于，在所述获取待校准电路板的图像之前，还包括：获取电路板的标准图像；从所述标准图像中选取N个特征点，并相应保存每个特征点的特征信息；所述特征信息包括特征点的图片和位置信息；N≥2；根据特征点的个数，配置匹配规则；所述匹配规则包括特征点的特征信息、图像转换类型和图像转换程度。
- 如权利要求2所述的图像校准方法，其特征在于，所述根据预设的匹配规则，调整所述待校准电路板的图像，使所述定位点与所述特征点的位置相同，具体包括：根据所述图像转换类型，在所述图像转换程度的范围内调整所述待校准电路板的图像，使每个定位点与其相应特征点的位置相同。
- 如权利要求2或3所述的图像校准方法，其特征在于，在所述根据特征点的个数，配置匹配规则之后，还包括：采用所述匹配规则，对电路板的采样图像进行校准测试；若校准测试不通过，则调整特征点的个数及其相应的匹配规则，并继续进行校准测试；若校准测试通过，则保存所述匹配规则。
- 如权利要求2或3所述的图像校准方法，其特征在于，所述图像转换类型包括旋转、仿射变换或透视变换中的至少一个；所述图像转换程度包括图像角度转换程度和图像比例转换程度。
- 一种图像校准装置，其特征在于，包括：图像获取模块，用于获取待校准电路板的图像；定位点识别模块，用于根据预存的特征点，从所述待校准电路板的图像中识别出定位点；所述特征点为标准图像中的图片区域；以及，图像调整模块，用于根据预设的匹配规则，调整所述待校准电路板的图像，使所述定位点与所述特征点的位置相同。
- 如权利要求6所述的图像校准装置，其特征在于，所述图像校准装置还包括：标准图像获取模块，用于获取电路板的标准图像；特征点选取模块，用于从所述标准图像中选取N个特征点，并相应保存每个特征点的特征信息；所述特征信息包括特征点的图片和位置信息；N≥2；以及，规则配置模块，用于根据特征点的个数，配置匹配规则；所述匹配规则包括特征点的特征信息、图像转换类型和图像转换程度。
- 如权利要求7所述的图像校准装置，其特征在于，所述图像调整模块具体用于根据所述图像转换类型，在所述图像转换程度的范围内调整所述待校准电路板的图像，使每个定位点与其相应特征点的位置相同。
- 如权利要求7或8所述的图像校准装置，其特征在于，所述图像校准装置还包括：校准测试模块，用于采用所述匹配规则，对电路板的采样图像进行校准测试；规则调整模块，用于若校准测试不通过，则调整特征点的个数及其相应的匹配规则，并继续进行校准测试；以及，保存模块，用于若校准测试通过，则保存所述匹配规则。
- 如权利要求7或8所述的图像校准装置，其特征在于，所述图像转换类型包括旋转、仿射变换或透视变换中的至少一个；所述图像转换程度包括图像角度转换程度和图像比例转换程度。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510770199.0A CN105427297B (zh) | 2015-11-12 | 2015-11-12 | 一种图像校准方法及装置 |
| CN201510770199.0 | 2015-11-12 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2017080283A1 true WO2017080283A1 (zh) | 2017-05-18 |
Family
ID=55505478
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2016/096893 WO2017080283A1 (zh) | 2015-11-12 | 2016-08-26 | 一种图像校准方法及装置 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN105427297B (zh) |
| WO (1) | WO2017080283A1 (zh) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105427297B (zh) * | 2015-11-12 | 2018-12-18 | 广州视源电子科技股份有限公司 | 一种图像校准方法及装置 |
| US10417738B2 (en) | 2017-01-05 | 2019-09-17 | Perfect Corp. | System and method for displaying graphical effects based on determined facial positions |
| CN109429473A (zh) * | 2017-08-28 | 2019-03-05 | 株洲中车时代电气股份有限公司 | 电路板中带极性电子元器件的自动检查方法及装置 |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1504742A (zh) * | 2002-11-28 | 2004-06-16 | 威光机械工程股份有限公司 | 印刷电路板上瑕疵组件的自动光学检测系统 |
| CN101334441A (zh) * | 2007-06-14 | 2008-12-31 | 雅马哈精密科技株式会社 | 印刷电路板的检查装置和检查方法 |
| CN102567724A (zh) * | 2010-12-11 | 2012-07-11 | 鸿富锦精密工业（深圳）有限公司 | 图像校正系统及方法 |
| CN105427297A (zh) * | 2015-11-12 | 2016-03-23 | 广州视源电子科技股份有限公司 | 一种图像校准方法及装置 |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102316242B (zh) * | 2010-07-06 | 2015-08-05 | 鸿富锦精密工业（深圳）有限公司 | 图像校正系统及其校正方法 |
| CN103020945B (zh) * | 2011-09-21 | 2016-02-10 | 中国科学院电子学研究所 | 一种多源传感器的遥感图像配准方法 |
| CN103226821B (zh) * | 2013-04-27 | 2015-07-01 | 山西大学 | 基于视差图像素分类校正优化的立体匹配方法 |
| TWI520099B (zh) * | 2014-02-19 | 2016-02-01 | 鈺創科技股份有限公司 | 影像擷取系統的校正方法 |
-
2015
- 2015-11-12 CN CN201510770199.0A patent/CN105427297B/zh active Active
-
2016
- 2016-08-26 WO PCT/CN2016/096893 patent/WO2017080283A1/zh active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1504742A (zh) * | 2002-11-28 | 2004-06-16 | 威光机械工程股份有限公司 | 印刷电路板上瑕疵组件的自动光学检测系统 |
| CN101334441A (zh) * | 2007-06-14 | 2008-12-31 | 雅马哈精密科技株式会社 | 印刷电路板的检查装置和检查方法 |
| CN102567724A (zh) * | 2010-12-11 | 2012-07-11 | 鸿富锦精密工业（深圳）有限公司 | 图像校正系统及方法 |
| CN105427297A (zh) * | 2015-11-12 | 2016-03-23 | 广州视源电子科技股份有限公司 | 一种图像校准方法及装置 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105427297B (zh) | 2018-12-18 |
| CN105427297A (zh) | 2016-03-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI585398B (zh) | Projection-type printed circuit board re-inspection system and methods, and marking the location of the defect | |
| WO2017080283A1 (zh) | 一种图像校准方法及装置 | |
| EP3864423A1 (en) | Compensating for reference misaligment during inspection of parts | |
| CN110858899A (zh) | 一种摄像机机芯光轴中心和视场角的测量方法及系统 | |
| WO2018068417A1 (zh) | 元件缺陷检测方法和系统 | |
| CN106651802B (zh) | 机器视觉焊锡定位检测方法 | |
| WO2017202114A1 (zh) | 确定用于检测的光照强度的方法和装置、及光学检测方法和装置 | |
| CN110146017A (zh) | 工业机器人重复定位精度测量方法 | |
| CN116503316A (zh) | 一种基于图像处理的芯片缺陷量测方法及系统 | |
| KR100208111B1 (ko) | 웨이퍼를 프로파일링하고 그 위에 다이의 위치를 설정하기 위한 방법 및 장치 | |
| CN104034259A (zh) | 一种影像测量仪校正方法 | |
| CN110470250B (zh) | 一种零件表面平面度的检测装置及检测方法 | |
| TW201928763A (zh) | 利用影像判斷電解電容極性的方法及系統 | |
| KR20160105082A (ko) | 기판 검사 장치 및 그 검사 방법 | |
| CN115881570A (zh) | 一种检测晶圆边缘形貌的方法及系统 | |
| US8055056B2 (en) | Method of detecting defects of patterns on a semiconductor substrate and apparatus for performing the same | |
| KR20120071842A (ko) | 마킹 위치 인식 장치 및 방법 | |
| KR102398139B1 (ko) | 로봇좌표계와 시편좌표계의 정렬장치 및 이를 이용한 좌표계 정렬 방법 | |
| KR20190079317A (ko) | 머신 비전 장치 및 머신 비전을 이용한 제품 검사 방법 | |
| CN110874837A (zh) | 一种基于局部特征分布的缺陷自动检测方法 | |
| KR20150049799A (ko) | 기판의 표면 검사 방법 및 이를 수행하기 위한 장치 | |
| TWI762364B (zh) | 晶錠評估方法 | |
| RU2813478C1 (ru) | Способ поверки или настройки тестера сверхбольших интегральных микросхем, манипулятор и плата для реализации способа | |
| KR102164675B1 (ko) | 극자외선 마스크의 결함 검출 방법 및 이를 수행하기 위한 장치 | |
| JP2903356B2 (ja) | Ｌｓｉチップの座標ズレ修正方法 |
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: 16863461 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: 16863461 Country of ref document: EP Kind code of ref document: A1 |