About the project

  • Applications: Vision Guided Robot
  • Industrial Sectors: Automotive Industry,Mechanical Industry
  • Technologies: 2D,3D
Specialvideo has developed a guidance system robot for taking bulk workpieces and folded sheets in a box (bin picking).

The three-dimensional laser scanning

The three-dimensional objects reconstruction is realized through a laser scanning, using the triangulation technique with a camera
A laser projects a bright line on the framed scene by the camera, so that the captured images contain the lines given by the intersection between the laser plane and the object surfaces.

Fig. 1 - Triangolazione laser
Fig. 1 - Laser triangulation

Knowing the laser plane equation and the camera position, through a triangulation process,we obtain the information necessary to determine the coordinates (x, y, z) belonging to the point P projected line on the object surface.
(Figures 1 and 2)

Fig. 2 - Triangolazione con angolo α
Fig. 2 - Triangulation with α angle
3A) telecamera verticale e laser inclinato
3A) vertical camera and inclined laser
3B) telecamera inclinata laser verticale
3B) vertical laser inclined camera
3C) telecamera e laser inclinati
3C) camera and laser both inclined

The possible laser-camera couple configurations are three. They can be selected according to the objectives to be achieved and the objects surface characteristics in the scene such as the reduction of sub-squares, the presence of reflective surfaces etc.
(Figure 3)

Fig. 4 - Scansione con il laser
Fig. 4 - Laser scan


The main advantages of this technique are:

  • substantial independence from the objects colors
  • independence from the object (texture) surface structure , unlike stereoscopic vision
  • the images obtained are "dense", that is, we obtain the three-dimensional information of the objects entire surface to exclusion of the sub-squares
  • resolution and appropriate precision of the robot guidance applications for large work areas

These advantages make the laser scanning one of the preferred methods to obtaining three-dimensional images in the industrial field.

System Features

The system function is to properly guide the robot assigning it the mechanical objects grasp to flat surface arranged randomly within a cassette.
The vision system, which arises as a solution to one of the most complex problems and most in demand in the artificial vision world in industrial field, must locate objects easily accessible by the picking robot.

The innovative pieces research algorithm is the result of years of dedicated research and development by Specialvideo in the 3D.

Fig. 5 - Calcolo dei piani di giacitura dei pezzi
Fig. 5 - Pieces calculation of the laying planes

The additional information provided by the third dimension are “simplified” up to calculate the pieces arrangement plane (Figure 5); then it can be applied to all 2D-established techniques such as pattern matching, quality control, dimensional measurements.

The images created by the vision system with laser triangulation are calibrated , ie the pixels are converted to millimeters according to a reference system shared with the Robot. In this way the coordinates exchange is correct and precise. The calibrated images are generated by a proprietary library of three-dimensional image processing where inside are implemented the most recent multidimensional calibration techniques.