Conversion of Halcon 3D depth map to 3D image

One 、 brief introduction

halcon Threshold segmentation of learning (threshold、binary_threshold、dyn_threshold、var_threshold、auto_threshold、fast_threshold、)_zsffuture The blog of -CSDN Blog _fast_threshold operator

Principle of turning depth map into point cloud - horsetail - Blog Garden

Depth images and 3D Point cloud rotation only involves camera intrinsic parameter matrix , among

• fx,fy They are lenses x,y The focal length of the direction （ The distance from the imaging plane to the lens ）
• cx,cy They are optical centers （ The projection of the lens center point on the imaging plane ） In the image coordinate system （ The origin is in the upper left corner of the image , Level is x, Vertical is y） The coordinates under .
• 

 

Two 、 Operator introduction

1、 Threshold segmentation

auto_threshold (AerialImage, Regions, 5)
* auto_threshold(Image : Regions : Sigma : )
* According to the histogram to determine the threshold, automatic global threshold segmentation , Operation principle , First of all , Calculate the gray histogram . second , Extract the minimum value from the histogram after Gaussian smoothing . Third , Threshold segmentation is performed according to the extracted minimum value .sigma The larger the extraction area, the less .

2、 Image conversion

* Convert the segmented region into a label image .
region_to_label (Regions, LabelImage, 'byte', Width, Height)
* region_to_label According to their index （1..n） Convert the input area into a label image , namely , The first area is drawn as a grayscale value 1, The second area is drawn as a gray value 2 etc. . Use only positive grayscale values .
 The third parameter  ：  *  about 'byte' Images , Index to get 256 Analog input of . The default is int2
*Width (input_control) The width of the image to be generated 
*Height (input_control)  The height of the image to be generated 

3、 Using polynomials will create a sloped gray surface  

gen_image_surface_first_order (X, 'real', 0, xishu, 0,  0, 0, Width, Height)
gen_image_surface_first_order (Y, 'real', -xishu, 0, 0, 0, 0, Width, Height)
*   Use first-order polynomials to create slanted gray surfaces . （ Output image , type , Vertical parameters , Horizontal parameters ,0 Order parameter , The row coordinates of the reference point , Column coordinates , The width and height of the image ）.

* ImageSurface(r,c) = Alpha(r - Row) +Beta(c - Column) +Gamma
*gen_image_surface_first_order( : ImageSurface : Type, Alpha, Beta, Gamma, Row, Column, Width, Height : )
*ImageSurface (output_object)  Newly created image matrix .
*Type (input_control) Pixel type   'byte'
*Alpha (input_control)  First order coefficient in the vertical direction 
*Beta (input_control) First order coefficient in horizontal direction 
*Gamma (input_control)　 Zero order coefficient 
*Row (input_control) Row coordinates of the surface reference point 
*Column (input_control) Column coordinates of the surface reference point 
*Width (input_control)  The width of the image 
*Height (input_control) 　 Height of the image