Halcon image rectification

  Image correction

**55 Represents the height of the maximum text 
text_line_orientation (RegionTrans, InputImage, 55, rad(-30), rad(30), OrientationAngle)
rotate_image (InputImage, ImageRotate, -OrientationAngle/rad(180)*180, 'constant')

 xld correct , I find this is not very clever .

get_image_size (InputImage, imageWidth, imageHeight)
binary_threshold (InputImage, Region2, 'max_separability', 'dark', UsedThreshold2)
fill_up (Region2, RegionFillUp)
erosion_circle (RegionFillUp, RegionErosion, 3.5)
gen_contour_region_xld (RegionErosion, Contours, 'border')
* Divide the profile into edges 
segment_contours_xld (Contours, ContoursSplit, 'lines', 5, 30, 2)
XCoordCorners := []
YCoordCorners := []
* Get the number of edges 
count_obj (ContoursSplit, Number)
* Store the starting position of each edge 
for index:=1 to Number by 1
   select_obj(ContoursSplit, ObjectCurrent, index)
   * Fit each edge 
   fit_line_contour_xld (ObjectCurrent, 'tukey', -1, 0, 5, 2, RowBegin, ColBegin, RowEnd, ColEnd, Nr, Nc, Dist)
   * Store the vertices of each edge x coordinate 
   tuple_concat (XCoordCorners, RowBegin, XCoordCorners)
   * Store the vertices of each edge y coordinate 
   tuple_concat (YCoordCorners, ColBegin, YCoordCorners)
endfor
*  The projection transformation establishes the correlation between the four feature points and the corrected coordinates 
XOff:= 100
YOff:= 100*imageHeight/imageWidth
hom_vector_to_proj_hom_mat2d (XCoordCorners, YCoordCorners, [1,1,1,1], [YOff,YOff,imageHeight-YOff,imageHeight-YOff], [XOff,imageWidth-XOff,imageWidth-XOff,XOff], [1,1,1,1], 'normalized_dlt', HomMat2D)
* Projection transformation 
projective_trans_image (InputImage, OuputImage, HomMat2D, 'bilinear', 'false', 'false')
*  Display the correction results 
dev_display (OuputImage)
return ()

Picture before correction :

  Corrected picture