Skip to content

Latest commit

 

History

History
988 lines (863 loc) · 36.4 KB

halcon.md

File metadata and controls

988 lines (863 loc) · 36.4 KB

第一章 机器视觉概述

filename

第二章 数字图像基础

filename

第三章 了解和熟悉HALCON

filename

第四章 图像预处理

filename

ROI选择

* 关闭窗口
dev_close_window ()
*读取图像
read_image (Clip, 'clip')
*把一个图像缩放到指定比例大小
zoom_image_factor (Clip, ImageZoomed, 0.5, 0.5, 'constant')
*获得图像尺寸
get_image_size (ImageZoomed, Width, Height)
*新建显示窗口，适应图像尺寸
dev_open_window (0, 0, Width, Height, 'black', WindowID)
dev_display (ImageZoomed)
*输入矩形长轴针对水平方向的角度
phi := 1.91
*输入矩形中心点的y值坐标、x值坐标
Row := 134
Column := 118
*选择ROI
gen_rectangle2 (Rectangle, Row, Column,phi, 19, 55)
*从原图中分割出ROI
reduce_domain (ImageZoomed, Rectangle, ImageReduced)
*显示分割后的ROI图像
dev_display (ImageReduced)

仿射变换

dev_close_window ()
read_image (Image, 'C:/Users/Dear/Desktop/《机器视觉案例》/案例原图/第四章/4-2.png')
get_image_size (Image, Width, Height)
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
dev_display (Image)
rgb1_to_gray (Image, GrayImage)
threshold (GrayImage, Regions, 200, 255)
area_center (Regions, Area, Row, Column)
hom_mat2d_identity (HomMat2DIdentity)
hom_mat2d_translate (HomMat2DIdentity, Height/2-Row, Width/2-Column, HomMat2DTranslate)
affine_trans_image (GrayImage, ImageAffinTrans, HomMat2DTranslate, 'constant', 'false')
hom_mat2d_rotate (HomMat2DIdentity, 3.14/2, Height/2, Width/2, HomMat2DRotate)
affine_trans_image (ImageAffinTrans, ImageAffinTrans1, HomMat2DRotate, 'constant', 'false')
hom_mat2d_scale (HomMat2DIdentity, 1.5, 1.5, Height/2, Width/2, HomMat2DScale)
affine_trans_image (ImageAffinTrans1, ImageAffinTrans2, HomMat2DScale, 'constant', 'false')

二维码校正

dev_close_window ()
read_image (Image, 'E:/《机器视觉案例》/案例原图/第四章/4-3.png')
dev_open_window_fit_image (Image, 0, 0, -1, -1, WindowHandle)
dev_set_color ('red')
dev_set_line_width (2)
X := [163,280,362,75]
Y := [125,120,361,340]
gen_cross_contour_xld (Crosses, X, Y, 6, 0.78)
dev_display (Image)
dev_display (Crosses)
hom_vector_to_proj_hom_mat2d (X, Y, [1,1,1,1], [75,360,360,75], [110,110,360,360], [1,1,1,1], 'normalized_dlt', HomMat2D)
projective_trans_image (Image, Image_rectified, HomMat2D, 'bilinear', 'false', 'false')

线性灰度变换

*关闭窗口
dev_close_window ()
*获取图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第四章/4-4 4-5.png')
*获取图像尺寸
get_image_size (Image, Width, Height)
*打开大小适应的窗口
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
*显示图像
dev_display (Image)
*图像灰度化
rgb1_to_gray (Image, GrayImage)
*图像取反
invert_image (GrayImage, ImageInvert)
*增强对比度
emphasize (GrayImage, ImageEmphasize, Width, Height, 2)
*缩放图像灰度值
scale_image (GrayImage, ImageScaled, 1, 100)
*分别显示ImageInvert、ImageEmphasize、ImageScaled
dev_display (ImageInvert)
dev_display (ImageEmphasize)
dev_display (ImageScaled)

直方图均衡化

*关闭窗口
dev_close_window ()
*获取图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第四章/4-4 4-5.png')
*获取图像大小
get_image_size (Image,Width, Height)
*打开与图像大小适应的窗口
dev_open_window (0, 0, Width, Height, 'black', WindowHandle1)
*图像灰度化
rgb1_to_gray (Image, GrayImage)
*直方图均衡化
equ_histo_image (GrayImage, ImageEquHisto)

均值滤波

*读取图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第四章/4-6 4-7.jpg')
*关闭窗口
dev_close_window ( )
*重新打开窗口
dev_open_window (0, 0, 512, 512, 'black', WindowHandle)
*显示图像，将图像灰度化
dev_display (Image)
rgb1_to_gray (Image, GrayImage)
*添加高斯噪声
gauss_distribution (20, Distribution)
add_noise_distribution (GrayImage, ImageNoise, Distribution)
*均值滤波处理
mean_image (ImageNoise, ImageMean, 5, 5)
*显示图像
dev_display (ImageMean)

中值滤波

*读取图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第四章/4-6 4-7.jpg')
rgb1_to_gray (Image, GrayImage)
*关闭窗口，重新打开窗口
dev_close_window ( )
dev_open_window (0, 0, 512, 512, 'black', WindowHandle)
*显示图片
dev_display (GrayImage)
*图像灰度化
*添加椒盐噪声
sp_distribution (3, 3, Distribution)
add_noise_distribution (GrayImage, ImageNoise, Distribution)
*中值滤波
median_image (ImageNoise, ImageMedian, 'circle', 3, 'mirrored')
*显示图像
dev_display (ImageMedian)

sobel算子

*读取图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第四章/4-8.png')
*边缘检测
sobel_amp (Image, EdgeAmplitude, 'sum_abs', 3)
*阈值分割
threshold (EdgeAmplitude, Region, 10, 255)
*提取边缘框架
skeleton (Region, Skeleton)
*显示图片
dev_display (Image)
*设置输出颜色为红色
dev_set_color ('red')
*显示边缘框架
dev_display (Skeleton)

拉普拉斯算子

*关闭窗口
dev_close_window ()
*获取图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第四章/4-9.png')
*获取图像尺寸
get_image_size (Image, Width, Height)
*打开适应图片大小的窗口
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
*显示图片
dev_display (Image)
*对图像进行拉普拉斯算子处理
laplace (Image, ImageLaplace, 'signed', 3, 'n_8_isotropic')
*显示处理后的图像
dev_display (ImageLaplace)

低通滤波处理

*读取带有椒盐噪声图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第四章/4-10.png')
*获得图像尺寸
get_image_size (Image, Width, Height)
*关闭窗口
dev_close_window ()
*打开适应图像大小的窗口
dev_open_window (0, 0, 512, 512, 'black', WindowHandle)
*显示图像
dev_display (Image)
*图像转灰度化
rgb1_to_gray (Image, GrayImage)
*获得一个低通滤波模型
gen_lowpass (ImageLowpass, 0.1, 'none', 'dc_center', Width, Height)
*对噪声图像进行傅里叶变换，得到频率图像
fft_generic (Image, ImageFFT, 'to_freq', -1, 'sqrt', 'dc_center', 'complex')
*对频率图像进行低通滤波
convol_fft (ImageFFT, ImageLowpass, ImageConvol)
*对频率图像进行傅里叶变换
fft_generic (ImageConvol, ImageFFT1, 'from_freq',1, 'sqrt', 'dc_center', 'complex')

高通滤波处理

*关闭窗口
dev_close_window ()
*读取图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第四章/4-11.png')
*图像灰度化
rgb1_to_gray (Image, GrayImage)
*获取图像大小
get_image_size (GrayImage, Width, Height)
*打开图像适应大小窗口
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
*得到高通滤波模型
gen_highpass (ImageHighpass, 0.1, 'n', 'dc_center', Width, Height)
*对图像进行傅里叶变换
fft_generic(GrayImage,ImageFFT,'to_freq',-1,'none','dc_center','complex')
*对频率图像进行高通滤波
convol_fft(ImageFFT,ImageHighpass,ImageConvol)
*对得到的频率图像进行傅里叶反变换
fft_generic(ImageConvol,ImageResult,'from_freq',1,'none','dc_center','byte')
*显示图像
dev_display (ImageResult)

第五章 图像分割

filename

直方图谷底确定阈值分割

read_image (Image, 'E:/《机器视觉案例》/案例原图/第五章/5-1.png')
get_image_size (Image, Width, Height)
dev_close_window ()
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
dev_set_color ('red')
*计算图像的灰度直方图
gray_histo (Image, Image, AbsoluteHisto, RelativeHisto)
histo_to_thresh (RelativeHisto, 8, MinThresh, MaxThresh)
dev_set_colored (12)
threshold (Image, Region, MinThresh[0], MaxThresh[0])
dev_display (Region)

基于直方图的自动阈值分割

*获取图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第五章/5-2.png')
*自动阈值分割
auto_threshold (Image, Regions, 5)
*显示分割区域
dev_display (Regions)

局部阈值分割算子dyn_threshold

*关闭窗口
dev_close_window ()
*获取图像
read_image (Image, 'photometric_stereo/embossed_01')
*获得图像尺寸
get_image_size (Image, Width, Height)
*打开适应图像大小的窗口
dev_open_window (0, 0, Width, Height, 'black', WindowHandle1)
*在图像上使用均值滤波器进行适当平滑
mean_image (Image, ImageMean, 59, 59)
*动态阈值分割，提取圆区域
dyn_threshold (Image, ImageMean, RegionDynThresh, 15, 'not_equal')
* 显示图像
dev_display (Image)
*显示提取区域
dev_display (RegionDynThresh)

Prewitt边缘提取实例

*读取图像
read_image (Image, 'fabrik')
*用prewitt算子进行边缘提取
prewitt_amp (Image, ImageEdgeAmp)
*进行阈值操作
threshold (ImageEdgeAmp, Region, 20, 255)
*骨骼化操作
skeleton (Region, Skeleton)
*显示图像
dev_display (Image)
*设置输出颜色为红色
dev_set_color ('red')
*显示骨骼图像
dev_display (Skeleton)

Kirsch边缘提取实例

*读取图像
read_image (Image, 'fabrik')
*用kirsch算子进行边缘检测
kirsch_amp (Image, ImageEdgeAmp)
*进行阈值操作
threshold (ImageEdgeAmp, Region, 70, 255)
*骨骼化操作
skeleton (Region, Skeleton)
*显示图像
dev_display (Image)
*设置输出颜色为红色
dev_set_color ('red')
*显示骨骼图像
dev_display (Skeleton)

Canny边缘提取分割实例

*获取图像
read_image (Image, 'fabrik')
*使用canny算法进行边缘提取
edges_image (Image, ImaAmp, ImaDir, 'lanser2', 0.5, 'nms', 12, 22)
*阈值分割
threshold (ImaAmp, Edges, 1, 255)
*骨骼化
skeleton (Edges, Skeleton)
*将骨骼化的区域转化为XLD
gen_contours_skeleton_xld (Skeleton, Contours, 1, 'filter')
*显示图像
dev_display (Image)
*设置6种输出颜色
dev_set_colored (6)
*显示XLD
dev_display (Contours)

亚像素级别的边缘提取实例

* 关闭窗口
dev_close_window ()
* 读取图像
read_image (Image, 'fabrik')
* 打开适应图像大小的窗口
dev_open_window_fit_image (Image, 0, 0, -1, -1, WindowHandle)
* 对图像进行亚像素区域提取
edges_sub_pix (Image, Edges, 'canny', 2, 12, 22)
* 放大图像用于详细的边缘检查
dev_set_part (160, 250, 210, 300)
dev_display (Image)
dev_display (Edges)

区域生长法实例

*读取图像
read_image (Image, 'fabrik')
*对图像进行均值处理，选用circle类型的中值滤波器
median_image (Image, ImageMedian, 'circle', 2, 'mirrored')
*使用regiongrowing算子寻找颜色相近的邻域
regiongrowing (ImageMedian, Regions, 1, 1, 2, 5000)
*对图像进行区域分割，提取满足各个条件的各个独立区域
shape_trans (Regions, Centers, 'inner_center')
connection (Centers, SingleCenters)
*计算出初步提取的区域的中心点坐标
area_center (SingleCenters, Area, Row, Column)
*以均值灰度图像为输入，进行区域增长计算，计算的起始坐标为上一步的各区域中心
regiongrowing_mean (ImageMedian, RegionsMean, Row, Column, 25, 100)

Hough变换图像分割实例

*读取图像
read_image (Image, 'fabrik')
*获得目标区域图像
rectangle1_domain (Image, ImageReduced, 170, 280, 310, 360)
* 用Sobel边缘检测算子提取边缘
sobel_dir (ImageReduced, EdgeAmplitude, EdgeDirection, 'sum_abs', 3)
*设置输出颜色为红色
dev_set_color ('red')
*阈值分割得到图像
threshold (EdgeAmplitude, Region, 55, 255)
* 截取图像
reduce_domain (EdgeDirection, Region, EdgeDirectionReduced)
* 进行Hough变换
hough_lines_dir (EdgeDirectionReduced, HoughImage, Lines, 4, 2, 'mean', 3, 25, 5, 5, 'true', Angle, Dist)
*将霍夫变换提取直线以普通形式描述的输入行存储为区域
gen_region_hline (LinesHNF, Angle, Dist)
*显示图像
dev_display (Image)
*设置输出颜色数目
dev_set_colored (12)
*设置输出填充方式为“轮廓”
dev_set_draw ('margin')
*显示LinesHNF
dev_display (LinesHNF)
*设置输出填充方式为“填充”
dev_set_draw ('fill')
*显示Lines
dev_display (Lines)

分水岭算法分割实例

*获取图像
read_image (Br2, 'particle')
*对单通道图像进行高斯平滑处理，去除噪声
gauss_filter (Br2, ImageGauss, 9)
*将图像颜色进行反转
invert_image (ImageGauss, ImageInvert)
*对高斯平滑后的图像进行分数岭处理与阈值分割，提取出盆地区域
watersheds (ImageInvert, Basins, Watersheds)
watersheds_threshold (ImageInvert, Basins1,30)

第六章 特征提取

filename

获取区域的面积和中线点特征实例

*获取图像
read_image (Image, 'fabrik')
*关闭窗口
dev_close_window ()
*打开窗口
dev_open_window (0, 0, 512, 512, 'black', WindowID)
*设置输出字体，14号字，Courier字体，粗体
set_display_font (WindowID, 14, 'mono', 'true', 'false')
*设置输出颜色
dev_set_colored (6)
*进行区域生长操作
regiongrowing (Image, Regions, 1, 1, 3, 200)
*显示区域
dev_display (Regions)
*计算所有不相连区域的面积和中心点坐标
area_center (Regions, Area, Row, Column)
center:=|Area|
*获取一个字符串的空间大小
get_string_extents (WindowID, 12345, Ascent, Descent, TxtWidth, TxtHeight)
*将面积计算结果以字符串形式显示在窗口中
for I := 0 to center-1 by 1
    disp_message (WindowID, Area[I], 'image', Row[I] - TxtHeight / 2, Column[I] - TxtWidth / 2, 'white', 'false')
endfor

计算封闭区域面积实例

*关闭窗口
dev_close_window ()
*读取图像
read_image (Image, 'rings_and_nuts')
*打开适应图像大小的窗口
dev_open_window_fit_image (Image, 0, 0, -1, -1, WindowHandle)
*设置输出颜色
dev_set_color ('red')
*设置输出字体，14号字，Courier字体，粗体
set_display_font (WindowHandle, 14, 'mono', 'true', 'false')
*显示图像
dev_display (Image)
*进行阈值操作
threshold (Image, Region, 128, 255)
* 计算区域中孔的面积
area_holes (Region, Area)
*将面积计算结果以字符串形式显示在窗口中
disp_message (WindowHandle, 'Size of enclosed area (holes): ' + Area + ' pixel', 'window', 12, 12, 'black', 'true')

利用Select_shape选择区域实例

*读取图像
read_image (Image, 'monkey')
*二值化
threshold (Image, Region, 128, 255)
*将非连通区域分割成一组区域的集合
connection (Region, ConnectedRegions)
*利用面积以及椭圆长轴与短轴的比值特征，将眼睛部分区域提取出来
select_shape (ConnectedRegions, SelectedRegions, ['area','anisometry'], 'and', [500,1], [2000,1.7])
*显示结果区域
dev_display (SelectedRegions)

使用inner circle算子计算图像区域内接圆

*读取图像
read_image (Image, 'fabrik')
*关闭窗口
dev_close_window ()
*打开窗口
dev_open_window (0, 0, 512, 512, 'black', WindowID)
*设置输出颜色为白色
dev_set_color ('white')
*设置图像模式为填充模式
dev_set_draw ('fill')
*使用regiongrowing算子寻找颜色相近的邻域
regiongrowing (Image, Regions, 1, 1, 3, 500)
*找出每个区域的最大内接圆以及内接圆的中心坐标和半径
inner_circle (Regions, Row, Column, Radius)
*设置输出颜色为红色
dev_set_color ('red')
*显示每个区域的内接圆
disp_circle (WindowID, Row, Column, Radius)

求图像中的外接矩形实例

*关闭窗口
dev_close_window ()
*读取图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第六章/6-5.png')
*获取图像尺寸
get_image_size (Image, Width, Height)
*打开适应图像大小的窗口
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
*灰度化
rgb1_to_gray (Image, GrayImage)
*使用阈值处理提取较暗的部分
threshold (GrayImage, Regions, 0, 253)
*求平行坐标的最小外接矩形
smallest_rectangle1 (Regions, Row1, Column1, Row2, Column2)
*填充模式为轮廓
dev_set_draw ('margin')
*根据矩形参数绘制矩形的轮廓
gen_rectangle1 (Rectangle, Row1, Column1, Row2, Column2)
*求区域的最小外接矩形
smallest_rectangle2 (Regions, Row, Column, Phi, Length1, Length2)
*根据矩形参数绘制矩形的轮廓
gen_rectangle2 (Rectangle1, Row, Column, Phi, Length1, Length2)

gray_features算子实例

* 读取图片
read_image (Image, 'monkey')
*对图像进行阈值处理，主要是将图像转化为区域
threshold (Image, Region, 1, 255)
*提取区域中最小灰度值
gray_features (Region, Image, 'min',MinDisp)
*提取区域中最大灰度值
gray_features (Region, Image, 'max', MaxDisp)

运用intensity算子计算区域的灰度值平均值和偏差

* 读取图像
read_image (Image, 'mreut')
*关闭窗口
dev_close_window ()
*获得图像尺寸
get_image_size (Image, Width, Height)
*打开适应图像大小的窗口
dev_open_window (0, 0, Width, Height, 'black', WindowID)
*显示图像
dev_display (Image)
*设置填充模式为轮廓
dev_set_line_width (5)
dev_set_draw ('margin')
* 创建两个矩形区域
gen_rectangle1 (Rectangle1, 350, 100, 450, 200)
gen_rectangle1 (Rectangle2, 100, 200, 200, 300)
*提取区域1中灰度值的平均值与偏差
intensity (Rectangle1, Image, Mean1, Deviation1)
*提取区域2中灰度值的平均值与偏差
intensity (Rectangle2, Image, Mean2, Deviation2)

使用min_max_gray算子提取区域中最大灰度以及最小灰度值

* 读取图像
read_image (Image, 'mreut')
*关闭窗口
dev_close_window ()
*获得图像尺寸
get_image_size (Image, Width, Height)
*打开适应图像大小的窗口
dev_open_window (0, 0, Width, Height, 'black', WindowID)
*显示图像
dev_display (Image)
*设置填充模式为轮廓
dev_set_line_width (5)
dev_set_draw ('margin')
* 创建两个矩形区域
gen_rectangle1 (Rectangle1, 350, 100, 450, 200)
gen_rectangle1 (Rectangle2, 100, 200, 200, 300)
*提取区域1中最大与最小灰度值
min_max_gray (Rectangle1, Image, 5, Min1, Max1, Range1)
*提取区域2中最大与最小灰度值
min_max_gray (Rectangle2, Image, 5, Min2, Max2, Range2)

运用area_center_gray 计算一幅灰度值图像的面积和中心实例

* 读取图像
read_image (Image, 'mreut')
* 创建矩形区域
gen_rectangle1 (Rectangle1, 350, 100, 450, 200)
* 计算区域内总灰度值以及中心的行、列坐标
area_center_gray (Rectangle1, Image, Area, Row, Column)

select_gray算子运算实例

*读取图像
read_image (Image, 'fabrik')
*关闭窗口
dev_close_window ()
*打开窗口
dev_open_window (0, 0, 512, 512, 'black', WindowID)
*设置输出颜色为白色
dev_set_color ('white')
*设置填充模式为填充
dev_set_draw ('fill')
*使用regiongrowing算子进行区域增长
regiongrowing (Image, Regions, 1, 1, 3, 10)
*设置输出颜色为红色
dev_set_color ('red')
*使用select_gray算子快速选择出灰度值在199-255碱的区域
select_gray (Regions, Image, SelectedRegions, 'mean', 'and',190, 250)

第七章 图像的形态学处理

filename

利用不同的腐蚀算子得到不同的腐蚀结果实例

dev_close_window ()
*获取图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第七章/7-1.png')
*获取图像的尺寸
get_image_size (Image, Width, Height)
*创建新的显示窗口（适应图像尺寸）
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
*显示图像
dev_display (Image)
*将图像转化为灰度图像
rgb1_to_gray (Image, GrayImage)
*将图像通过阈值处理转化为二值化图像
threshold (GrayImage, Regions, 134, 239)
*使用半径为11的圆形结构腐蚀得到区域
erosion_circle (Regions, RegionErosion, 11)
*使用长宽均为11的矩形结构元素腐蚀得到区域
erosion_rectangle1 ( Regions, RegionErosion1, 11, 11)
*生成短轴11，长轴13的椭圆形区域，作为结构元素
gen_ellipse (Ellipse, 100, 100, 0, 13, 11)
*使用生成的椭圆形结构元素腐蚀得到区域
erosion1 (Regions, Ellipse, RegionErosion2, 1)
*使用生成的椭圆形结构元素腐蚀得到区域（可设置参考点）
erosion2 (Regions, Ellipse, RegionErosion3, 0, 0, 1)

利用不同的膨胀算子得到不同的膨胀结果实例

dev_close_window ()
*获取图像
read_image (Image71, 'D:/《机器视觉》图片/第七章/7-1.png')
get_image_size (Image71, Width, Height)
*创建新的显示窗口（适应图像尺寸）
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
*显示图像
dev_display (Image71)
*将图像转化为灰度图像
rgb1_to_gray (Image71, GrayImage)
*将图像通过阈值处理转化为二值化图像
threshold (GrayImage, Regions, 134, 239)
*使用半径为11的圆形结构膨胀得到区域
dilation_circle (Regions, RegionDilation, 11)
*使用长为13，宽为11的矩形结构元素膨胀得到区域
dilation_rectangle1 (Regions, RegionDilation1, 11, 13)
*生成短轴11，长轴13的椭圆形区域，作为结构元素
gen_ellipse (Ellipse, 100, 100, 0, 13, 11)
*使用生成的椭圆形结构元素膨胀得到区域
dilation1 (Regions, Ellipse, RegionDilation2, 1)
*使用生成的椭圆形结构元素膨胀得到区域（可设置参考点）
dilation2 (Regions, Ellipse, RegionDilation3, 0, 0, 1)

HALCON闭运算实例

read_image (Image, 'E:/《机器视觉案例》/案例原图/第七章/7-3 闭运算.png')
dev_close_window ()
rgb1_to_gray (Image, GrayImage)
get_image_size (GrayImage, Width, Height)
dev_open_window (0, 0, Width / 2, Height / 2, 'black', WindowID)
dev_display (GrayImage)
threshold (GrayImage, Regions1, 25, 189)
closing_circle (Regions1, RegionClosing, 7)
closing_rectangle1 (Regions1, RegionClosing1, 9, 9)
gen_ellipse (Ellipse, 200, 200, 0, 5, 7)
closing (Regions1, Ellipse, RegionClosing2)
dev_display (RegionClosing)
dev_display (RegionClosing1)
dev_display (RegionClosing2)


dev_close_window ()
read_image (Image, 'E:/《机器视觉案例》/案例原图/第七章/7-3开运算.png')
rgb1_to_gray (Image, GrayImage)
get_image_size (GrayImage, Width, Height)
dev_open_window (0, 0, Width / 2, Height / 2, 'black', WindowID)
dev_display (GrayImage)
threshold (GrayImage, Regions1, 25, 189)
opening_circle (Regions1, RegionOpening,11)
opening_rectangle1 (Regions1, RegionOpening1,11, 11)
gen_ellipse (Ellipse, 200, 200, 0, 11, 13)
opening (Regions1, Ellipse, RegionOpening2)
dev_display (RegionOpening)
dev_display (RegionOpening1)
dev_display (RegionOpening2)

灰度图像的腐蚀与膨胀、开运算与闭运算操作实例

*关闭窗口
dev_close_window ()
*获取图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第七章/7-4.png')
*获取图像尺寸大小
get_image_size (Image, Width, Height)
*打开与图像适应大小的窗口
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
*显示图像
dev_display (Image)
*对灰度图像进行腐蚀操作
gray_erosion_shape (Image, ImageMin, 5, 5, 'octagon')
*对灰度图像进行膨胀操作
gray_dilation_shape (Image, ImageMax, 5, 5, 'octagon')
*对灰度图像进行开运算操作
gray_opening_shape (Image, ImageOpening, 5, 5, 'octagon')
*对灰度图像进行闭运算操作
gray_closing_shape (Image, ImageClosing, 5, 5, 'octagon')

灰度图像的顶帽、底帽运算

*关闭窗口
dev_close_window ()
*获取图像
read_image (Image, 'E:/《机器视觉案例》/案例原图/第七章/7-5.png')
*获取图片图像
get_image_size (Image, Width, Height)
*打开适应图像大小的窗口
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
*显示图像
dev_display (Image)
*图像灰度化
rgb1_to_gray (Image, GrayImage)
*生成结构半径为5的圆形结构
gen_disc_se (SE, 'byte', 21, 21, 0)
*对图像进行顶帽运算处理
gray_tophat (GrayImage, SE, ImageTopHat)
*对图像进行底帽运算处理
gray_bothat (GrayImage, SE, ImageBotHat)

边界提取实例

*关闭窗口
dev_close_window ()
*获取图像
read_image (Image76, 'E:/《机器视觉案例》/案例原图/第七章/7-6.png')
*获取图像大小
get_image_size (Image76, Width, Height)
*打开适应图像大小的窗口
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
*显示图像
dev_display (Image76)
*将图像二值化
threshold (Image76, Regions, 76, 255)
*边界提取
boundary (Regions, RegionBorder, 'inner')

孔洞填充实例

*获取图像
read_image (Aegyt1, 'egypt1')
*将图像二值化
threshold (Aegyt1, Region, 0, 140)
*填充孔洞
fill_up (Region, RegionFillUp)

骨架提取实例

read_image (Image, 'fabrik')
dev_close_window ()
get_image_size (Image, Width, Height)
dev_open_window (0, 0, Width, Height, 'black', WindowID)
dev_display (Image)
*边缘检测
edges_image (Image, ImaAmp, ImaDir, 'lanser2', 0.5, 'nms', 8, 16)
*阈值分割
threshold (ImaAmp, Region, 8, 255)
*提取骨架
skeleton (Region, Skeleton)

第八章 图像模板匹配

filename

基于像素灰度值的模板匹配

read_image (Image, 'smd/smd_on_chip_05')
get_image_size (Image, Width, Height)
dev_close_window ()
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
set_display_font (WindowHandle, 16, 'mono', 'true', 'false')
dev_set_color ('green')
dev_set_draw ('margin')
gen_rectangle1 (Rectangle, 175, 156, 440, 460)
area_center (Rectangle, Area, RowRef, ColumnRef)
reduce_domain (Image, Rectangle, ImageReduced)
create_ncc_model (ImageReduced, 'auto', 0, 0, 'auto', 'use_polarity', ModelID)
dev_display (Image)
dev_display (Rectangle)
disp_continue_message (WindowHandle, 'black', 'true')
for J := 1 to 11 by 1
    read_image (Image, 'smd/smd_on_chip_' + J$'02')
    find_ncc_model (Image, ModelID, 0, 0, 0.5, 1, 0.5, 'true', 0, Row, Column, Angle, Score)
    vector_angle_to_rigid (RowRef, ColumnRef, 0, Row, Column, 0, HomMat2D)
    affine_trans_region (Rectangle, RegionAffineTrans, HomMat2D, 'nearest_neighbor')
    dev_display (Image)
    dev_display (RegionAffineTrans)
    if (J < 11)
        disp_continue_message (WindowHandle, 'black', 'true')
    endif
    stop ()
endfor
clear_ncc_model (ModelID)

基于形状的模板匹配

*读取图像
read_image (Image, 'green-dot')
get_image_size (Image, Width, Height)
dev_close_window ()
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)
dev_set_color ('red')
dev_set_line_width (5)
dev_display (Image)
*提取模板
threshold (Image, Region, 0, 128)
connection (Region, ConnectedRegions)
select_shape (ConnectedRegions, SelectedRegions, 'area', 'and', 10000, 20000)
fill_up (SelectedRegions, RegionFillUp)
dilation_circle (RegionFillUp, RegionDilation, 5.5)
reduce_domain (Image, RegionDilation, ImageReduced)
*创建模版
create_scaled_shape_model (ImageReduced, 5, rad(-45), rad(90), 'auto', 0.8, 1.0, 'auto', 'none', 'ignore_global_polarity', 40, 10, ModelID)
*提取模版轮廓
get_shape_model_contours (Model, ModelID, 1)
*显示模板轮廓到模版位置
area_center (RegionFillUp, Area, RowRef, ColumnRef)
vector_angle_to_rigid (0, 0, 0, RowRef, ColumnRef, 0, HomMat2D)
affine_trans_contour_xld (Model, ModelTrans, HomMat2D)
dev_display (Image)
dev_display (ModelTrans)
*读取目标图片
read_image (ImageSearch, 'green-dots')
dev_display (ImageSearch)
*匹配模版
find_scaled_shape_model (ImageSearch, ModelID, rad(-45), rad(90), 0.8, 1.0, 0.5, 0, 0.5, 'least_squares', 5, 0.8, Row, Column, Angle, Scale, Score)
*循环显示匹配出来的模版区域轮廓
for I := 0 to |Score| - 1 by 1
    hom_mat2d_identity (HomMat2DIdentity)
    hom_mat2d_translate (HomMat2DIdentity, Row[I], Column[I], HomMat2DTranslate)
    hom_mat2d_rotate (HomMat2DTranslate, Angle[I], Row[I], Column[I], HomMat2DRotate)
    hom_mat2d_scale (HomMat2DRotate, Scale[I], Scale[I], Row[I], Column[I], HomMat2DScale)
    affine_trans_contour_xld (Model, ModelTrans, HomMat2DScale)
    dev_display (ModelTrans)
endfor
*清除模版
clear_shape_model (ModelID)

图像金字塔应用

*使用采集助手读取图像
open_framegrabber ('File', 1, 1, 0, 0, 0, 0, 'default', -1, 'default', -1, 'default', 'rings/rings.seq', 'default', -1, 1, FGHandle)
grab_image (ModelImage, FGHandle)
get_image_pointer1 (ModelImage, Pointer, Type, Width, Height)
dev_close_window ()
dev_open_window (0, 0, Width, Height, 'white', WindowHandle)
dev_display (ModelImage)
* 设置显示条件
dev_set_color ('red')
dev_set_draw ('margin')
dev_set_line_width (5)
*生成模版的圆形区域并截取模版
gen_circle (ModelROI, 251, 196, 103)
reduce_domain (ModelImage, ModelROI, ImageROI)
*创建基于金字塔层数的模型表示
inspect_shape_model (ImageROI, ShapeModelImage, ShapeModelRegion, 5, 30)
dev_display (ShapeModelRegion)
*创建基于形状的模版
create_shape_model (ImageROI, 'auto', 0, rad(360), 'auto', 'none', 'use_polarity', 30, 10, ModelID)
get_shape_model_contours (ShapeModel, ModelID, 1)
*循环识别目标图像
for i := 1 to 7 by 1
    *读取图像
    grab_image (SearchImage, FGHandle)
    dev_display (SearchImage)
    *匹配模版
    find_shape_model (SearchImage, ModelID, 0, rad(360), 0.6, 0, 0.55, 'least_squares', 0, 0.8, RowCheck, ColumnCheck, AngleCheck, Score)
    *显示匹配结果
    for j := 0 to |Score| - 1 by 1
        vector_angle_to_rigid (0, 0, 0, RowCheck[j], ColumnCheck[j], AngleCheck[j], MovementOfObject)
        affine_trans_contour_xld (ShapeModel, ModelAtNewPosition, MovementOfObject)
        dev_set_color ('blue')
        dev_display (ModelAtNewPosition)
        dev_set_color ('red')
        affine_trans_pixel (MovementOfObject, -120, 0, RowArrowHead, ColumnArrowHead)
        disp_arrow (WindowHandle, RowCheck[j], ColumnCheck[j], RowArrowHead, ColumnArrowHead, 2)
    endfor
endfor
*清除模版和关闭采集助手
clear_shape_model (ModelID)
close_framegrabber (FGHandle)

第九章 3D立体视觉

filename

第十章 HALCON相关实例和算法

filename

卡号识别系统案例

dev_clear_window ()

read_image (Image1, '9-3_0.bmp')
read_image (Image2, '9-1_1.bmp')

get_image_size (Image1, Width, Height)
dev_close_window ()
dev_open_window (0, 0, Width, Height, 'black', WindowHandle)

dev_display (Image1)
gen_rectangle1 (ROI_1, 13.5, 106.232, 61.5, 355.273)
reduce_domain (Image1, ROI_1, ImageReduced1)
area_center (ROI_1, Area1, Row1, Column1)

dev_display (Image2)
gen_rectangle1 (ROI_2, 403.5, 133.5, 547.5, 188.5)
reduce_domain (Image2, ROI_2, ImageReduced2)
area_center (ROI_2, Area2, Row2, Column2)

create_shape_model (ImageReduced1, 'auto', 0, rad(360), 'auto', 'auto', 'use_polarity', 'auto', 'auto', ModelID1)
create_shape_model (ImageReduced2, 'auto', 0, rad(360), 'auto', 'auto', 'use_polarity', 'auto', 'auto', ModelID2)

ImageFiles := []
ImageFiles[0] := '9-1_1.bmp'
ImageFiles[1] := '9-3_0.bmp'
for Index := 0 to |ImageFiles| - 1 by 1
    read_image (Image, ImageFiles[Index])
    find_shape_model (Image, ModelID1, 0, rad(360), 0.5, 1, 0.5, 'least_squares', 0, 0.9, Row11, Column11, Angle11, Score1)
    find_shape_model (Image, ModelID2, 0, rad(360), 0.5, 1, 0.5, 'least_squares', 0, 0.9, Row21, Column21, Angle21, Score2)
    if(|Score1|==1)
        vector_angle_to_rigid (Row11, Column11, Angle11, Row1, Column1, rad(0), HomMat2D)
        affine_trans_image (Image, ImageAffinTrans1, HomMat2D, 'constant', 'false')
        dev_display (ImageAffinTrans1)
        rgb1_to_gray (ImageAffinTrans1, GrayImage1)
        threshold (GrayImage1, Regions1, 9, 70)
        connection (Regions1, ConnectedRegions1)
        select_shape (ConnectedRegions1, SelectedRegions1, ['area','row','column'], 'and', [81.66,234.92,66.58], [207.29,255.03,483.67])
        union1 (SelectedRegions1, RegionUnion1)
        shape_trans (RegionUnion1, RegionTrans1, 'rectangle1')
        orientation_region (RegionTrans1, Phi1)
        area_center (RegionTrans1, Area12, Row12, Column12)
        vector_angle_to_rigid (Row12, Column12, Phi1, Row12, Column12, rad(0), HomMat2D2)
        reduce_domain (GrayImage1, RegionTrans1, ImageReduced1)
        rgb1_to_gray (ImageReduced1, GrayImage2)
        threshold (GrayImage2, Regions2, 11, 104)
        connection (Regions2, ConnectedRegions1)
        sort_region (ConnectedRegions1, SortedRegions1, 'character', 'true', 'column')
        read_ocr_class_mlp ('Industrial_0-9_NoRej.omc', OCRHandle1)
        do_ocr_multi_class_mlp (SortedRegions1, GrayImage2, OCRHandle1, Class1, Confidence1)
        disp_message (WindowHandle, Class1, 'window', 0, 0, 'black', 'true')
    endif
    if(|Score2|==1)
        vector_angle_to_rigid (Row21, Column21, Angle21, Row2, Column2,rad(0), HomMat2D3)
        affine_trans_image (Image, ImageAffinTrans2, HomMat2D3, 'constant', 'false')
        dev_display (ImageAffinTrans2)
        gen_rectangle1 (ROI_0, 449.5, 186.5, 467.5, 354.5)
        reduce_domain (ImageAffinTrans2, ROI_0, ImageReduced3)
        rgb1_to_gray (ImageReduced3, GrayImage3)
        threshold (GrayImage3, Regions3, 10, 57)
        connection (Regions3, ConnectedRegions2)
        partition_dynamic (ConnectedRegions2, Partitioned, 7, 0)
        select_shape (ConnectedRegions2, SelectedRegions2, 'column', 'and', 184.76, 201.69)
        select_shape (Partitioned, SelectedRegions3, 'column', 'and', 201.69, 343.13)
        select_shape (Partitioned, SelectedRegions4, 'column', 'and', 344.12, 356.08)
        sort_region (SelectedRegions3, SortedRegions2, 'character', 'true', 'column')
        read_ocr_class_mlp ('Industrial_A-Z+_NoRej.omc', OCRHandle2)
        read_ocr_class_mlp ('Industrial_0-9_NoRej.omc', OCRHandle3)
        read_ocr_class_mlp ('Industrial_0-9A-Z_NoRej.omc', OCRHandle4)
        do_ocr_multi_class_mlp (SelectedRegions2, GrayImage3, OCRHandle2, Class2, Confidence5)
        do_ocr_multi_class_mlp (SortedRegions2, GrayImage3, OCRHandle3, Class3, Confidence6)
        do_ocr_multi_class_mlp (SelectedRegions4, GrayImage3, OCRHandle4, Class4, Confidence7)
        disp_message (WindowHandle, [Class2,Class3,Class4], 'window', 0, 0, 'black', 'true')
    endif
endfor