公开(公告)号：US20110286659A1

公开(公告)日：2011-11-24

申请号：US13091218

申请日：2011-04-21

申请人： Kazuhito Saeki

发明人： Kazuhito Saeki

IPC分类号： G06K9/00

CPC分类号： G06T7/001 ,  G06T2207/10024 ,  G06T2207/30164

摘要： A plurality of first multivalued images of non-defective items picked up by an image pickup device are obtained, and a plurality of characteristic amounts at least including a pixel value of a color component and edge intensities in two different directions are extracted for each pixel in the obtained first multivalued images. Selection of any one of the plurality of extracted characteristic amounts is received, and the characteristic amount of which selection is received is extracted, whereby a distribution range for determining the non-defective item is calculated. A second multivalued image of a determination target object is obtained, and the characteristic amount of which selection is received is extracted for each pixel in the obtained second multivalued image, whereby determination is made as to whether the characteristic amount is included in the distribution range corresponding to the extracted characteristic amount.

摘要翻译： 获得由图像拾取装置拾取的无缺陷品的多个第一多值图像，并且针对每个像素提取至少包括颜色分量的像素值和两个不同方向的边缘强度的多个特征量 获得的第一多值图像。 接收多个提取的特征量中的任何一个的选择，并且提取接收到选择的特征量，从而计算用于确定不缺陷项目的分布范围。 获得确定目标对象的第二多值图像，并且对于获得的第二多值图像中的每个像素提取接收到选择的特征量，由此确定特征量是否包括在对应的分布范围中 提取特征量。

公开(公告)号：US06535630B1

公开(公告)日：2003-03-18

申请号：US09375464

申请日：1999-08-17

申请人： Kazuhito Saeki

发明人： Kazuhito Saeki

IPC分类号： G06K900

CPC分类号： G06K9/033 ,  G06F3/13 ,  G06K9/2054 ,  G09G5/04

摘要： In an image display method, each pixel belonging to one or more extraction areas with the lightness of the original image containing the pixel being contained in all lightness ranges set for the extraction areas is displayed in a first display color, each pixel belonging to one or more extraction areas with the lightness of the original image containing the pixel being out of all lightness ranges set for the extraction areas is displayed in a second display color, each pixel belonging to more than one extraction area with the lightness of the original image containing the pixel being contained in any of the lightness ranges set for the extraction areas and out of any of the lightness ranges is displayed in a third display color, and each pixel not belonging to any extraction area is displayed in the display color of the original image containing the pixel.

摘要翻译： 在图像显示方法中，以第一显示颜色显示属于一个或多个提取区域的，具有包含针对提取区域设置的所有亮度范围内的像素的原始图像的亮度的一个或多个提取区域，每个像素属于一个或多个 以第二显示颜色显示包含针对提取区域设置的所有亮度范围之外的像素的原始图像的亮度的更多提取区域，每个像素属于多于一个提取区域，原始图像的亮度包含 以提取区域和任何亮度范围设置的任何亮度范围中包含的像素以第三显示颜色显示，并且不属于任何提取区域的每个像素以包含的原始图像的显示颜色显示 像素。

公开(公告)号：US08594416B2

公开(公告)日：2013-11-26

申请号：US13091218

申请日：2011-04-21

申请人： Kazuhito Saeki

发明人： Kazuhito Saeki

IPC分类号： G06K9/00

CPC分类号： G06T7/001 ,  G06T2207/10024 ,  G06T2207/30164

摘要： A plurality of first multivalued images of non-defective items picked up by an image pickup device are obtained, and a plurality of characteristic amounts at least including a pixel value of a color component and edge intensities in two different directions are extracted for each pixel in the obtained first multivalued images. Selection of any one of the plurality of extracted characteristic amounts is received, and the characteristic amount of which selection is received is extracted, whereby a distribution range for determining the non-defective item is calculated. A second multivalued image of a determination target object is obtained, and the characteristic amount of which selection is received is extracted for each pixel in the obtained second multivalued image, whereby determination is made as to whether the characteristic amount is included in the distribution range corresponding to the extracted characteristic amount.

摘要翻译： 获得由图像拾取装置拾取的无缺陷品的多个第一多值图像，并且针对每个像素提取至少包括颜色分量的像素值和两个不同方向的边缘强度的多个特征量 获得的第一多值图像。 接收多个提取的特征量中的任何一个的选择，并且提取接收到选择的特征量，从而计算用于确定不缺陷项目的分布范围。 获得确定目标对象的第二多值图像，并且对于获得的第二多值图像中的每个像素提取接收到选择的特征量，由此确定特征量是否包括在对应的分布范围中 提取特征量。

公开(公告)号：US20110285843A1

公开(公告)日：2011-11-24

申请号：US13091252

申请日：2011-04-21

申请人： Kazuhito Saeki

发明人： Kazuhito Saeki

IPC分类号： H04N5/228 ,  H04N7/18

CPC分类号： G06T7/001 ,  G06T2207/10024

摘要： The invention provides an image processing apparatus, capable of detecting even a defect that exists in proximity to a contour line with high accuracy and determining a non-defective item with high accuracy. Edge intensities in two different directions are calculated for each pixel in the obtained first multivalued images, and a mean value of the edge intensities is calculated for each pixel in the first multivalued images. An intercorrelation distribution region of the edge intensities is calculated for each pixel in first multivalued images with the calculated mean value being the center. Edge intensities are calculated for each pixel in a second multivalued image of a determination target object, and determination is made as to whether the calculated edge intensities for each pixel in the second multivalued image are included in the calculated intercorrelation distribution region of the edge intensities for each pixel in the first multivalued images.

摘要翻译： 本发明提供一种图像处理装置，能够高精度地检测出与轮廓线附近存在的缺陷，能够高精度地确定无缺陷物品。 针对所获得的第一多值图像中的每个像素计算两个不同方向的边缘强度，并计算第一多值图像中的每个像素的边缘强度的平均值。 对于计算出的平均值为中心的第一多值图像中的每个像素计算边缘强度的相互关系分布区域。 针对确定目标对象的第二多值图像中的每个像素计算边缘强度，并且确定所计算的第二多值图像中的每个像素的边缘强度是否包括在所计算的边缘强度的互相关分布区域中， 第一个多值图像中的每个像素。

公开(公告)号：US07403218B2

公开(公告)日：2008-07-22

申请号：US10689756

申请日：2003-10-21

申请人： Kazuhito Saeki

发明人： Kazuhito Saeki

IPC分类号： H04N7/18

CPC分类号： G06T7/0004 ,  G06T2207/30164

摘要： An image processing apparatus has a trigger receiving section for receiving an external trigger and a trigger generation section including a timer. When a trigger from the outside is inputted, in synchronization with this trigger, the timer of the trigger generation section operates and a predetermined number of internal triggers are generated at predetermined intervals. After each image pickup data picked up by the external trigger and the internal triggers is processed by an image processing section statistical processing is internally performed by a statistical processing section and the maximum value, the minimum value, the average value, etc. of variations in the workpiece are calculated and the results are outputted.

摘要翻译： 图像处理装置具有用于接收外部触发的触发接收部分和包括定时器的触发器生成部分。 当从外部触发输入时，与该触发同步，触发生成部的定时器动作，以规定间隔生成规定数量的内部触发。 在由外部触发拾取的每个图像拾取数据和内部触发器由图像处理部分处理之后，统计处理在内部由统计处理部分执行，并且最大值，最小值，平均值等 计算工件并输出结果。

公开(公告)号：US06549655B1

公开(公告)日：2003-04-15

申请号：US09372756

申请日：1999-08-11

申请人： Yasuhisa Ikushima ,  Kazuhito Saeki

发明人： Yasuhisa Ikushima ,  Kazuhito Saeki

IPC分类号： G06K900

CPC分类号： G06K9/38 ,  G06T7/0004 ,  G06T7/11 ,  G06T7/136 ,  G06T2207/30164

摘要： A binarization method of binarizing an image by extracting lightness (brightness, density) as a feature amount from the image. When a pixel is selected in an image, a sensitivity is added to and/or subtracted from the value concerning the Y value of the selected pixel to set a threshold value range. Next, when another pixel is selected, the sensitivity is added to or subtracted from the value concerning the Y value of the selected pixel and a new threshold value range is set containing the calculation result and the already setup threshold value range. The pixel with the value concerning the Y value of any pixel in the image within the threshold value range is extracted as the same brightness as the selected pixel and the extraction result is displayed.

摘要翻译： 通过从图像中提取亮度（亮度，密度）作为特征量来二值化图像的二值化方法。 当在图像中选择像素时，将灵敏度添加到与所选像素的Y值相关的值中和/或从相关值中减去以设置阈值范围。 接下来，当选择另一个像素时，将灵敏度添加到与所选择的像素的Y值相关的值中或从相关值中减去，并且设置包含计算结果和已建立阈值范围的新阈值范围。 将与阈值范围内的图像中的任何像素的Y值相关的值的像素提取为与所选择的像素相同的亮度，并显示提取结果。

公开(公告)号：US07702158B2

公开(公告)日：2010-04-20

申请号：US10689755

申请日：2003-10-21

申请人： Kazuhito Saeki

发明人： Kazuhito Saeki

IPC分类号： G06K9/56 ,  G06K9/40

CPC分类号： G06T7/12 ,  G06T2207/30164

摘要： Plural window elements are set in one edge window, and edge detection is performed every each the window element, and calculation and determination about its result are made inside an image processing apparatus. A user can freely set a width of the window element and a distance between the adjacent window elements.

摘要翻译： 多个窗口元素设置在一个边缘窗口中，并且每个窗口元素执行边缘检测，并且在图像处理设备内进行关于其结果的计算和确定。 用户可以自由地设置窗口元素的宽度和相邻窗口元素之间的距离。

公开(公告)号：US08866903B2

公开(公告)日：2014-10-21

申请号：US13091252

申请日：2011-04-21

申请人： Kazuhito Saeki

发明人： Kazuhito Saeki

IPC分类号： H04N9/47 ,  H04N7/18 ,  G06T7/00

CPC分类号： G06T7/001 ,  G06T2207/10024

摘要： The invention provides an image processing apparatus, capable of detecting even a defect that exists in proximity to a contour line with high accuracy and determining a non-defective item with high accuracy. Edge intensities in two different directions are calculated for each pixel in the obtained first multivalued images, and a mean value of the edge intensities is calculated for each pixel in the first multivalued images. An intercorrelation distribution region of the edge intensities is calculated for each pixel in first multivalued images with the calculated mean value being the center. Edge intensities are calculated for each pixel in a second multivalued image of a determination target object, and determination is made as to whether the calculated edge intensities for each pixel in the second multivalued image are included in the calculated intercorrelation distribution region of the edge intensities for each pixel in the first multivalued images.

摘要翻译： 本发明提供一种图像处理装置，能够高精度地检测出与轮廓线附近存在的缺陷，能够高精度地确定无缺陷物品。 针对所获得的第一多值图像中的每个像素计算两个不同方向的边缘强度，并计算第一多值图像中的每个像素的边缘强度的平均值。 对于计算出的平均值为中心的第一多值图像中的每个像素计算边缘强度的相互关系分布区域。 针对确定目标对象的第二多值图像中的每个像素计算边缘强度，并且确定所计算的第二多值图像中的每个像素的边缘强度是否包括在所计算的边缘强度的互相关分布区域中， 第一个多值图像中的每个像素。

公开(公告)号：US07274812B2

公开(公告)日：2007-09-25

申请号：US10649308

申请日：2003-08-27

申请人： Kazuhito Saeki

发明人： Kazuhito Saeki

IPC分类号： G06K9/00 ,  H04N7/18

CPC分类号： G06T7/0004 ,  G06T7/0006 ,  G06T7/254 ,  Y10T29/49131

摘要： In an image processing device capable of collectively estimating an object based upon results of plural times of inspection, regions (1(1)) and (1(2)) to be photographed are defined on opposite end portions of an object (w). When an image of the left end portion of the object (w) is captured in response to a first trigger, the image processing device acquires corresponding image data, and the length (x1) of a projection extending in the x-direction at the left end of the object (w). When an image of the right end portion of the object (2) is next captured in response to a second trigger, the image processing device captures corresponding image data and measures the length (x2) of a projection extending in the x-direction at the right end of the work (w). Thereafter, the device finds the sum (x1+x2) of the lengths (x1, x2) inside, and outputs a result externally.

摘要翻译： 在能够基于多次检查的结果来集体估计物体的图像处理装置中，在物体（w）的相对端部限定要拍摄的区域（1（1））和（1（2））。 当响应于第一触发而捕获对象（w）的左端部分的图像时，图像处理装置获取对应的图像数据，并且在x方向上延伸的投影的长度（x 1） 对象的左端（w）。 当响应于第二触发而接下来捕获对象（2）的右端部分的图像时，图像处理装置捕获相应的图像数据并测量在x方向上延伸的突起的长度（x 2） 工作的右端（w）。 此后，设备找到内部长度（x 1，x 2）的和（x 1 + x 2），并从外部输出结果。