Academia.eduAcademia.edu

Figure 8 – uploaded by Luis Avila

See full PDFdownloadDownload figure
Figure 8: Some of the initial populations for the ge- netic algorithm.

Figure 8 Some of the initial populations for the ge- netic algorithm.

Related Figures (15)

Figure 1: 6 DOF robot manipulator with controller, and the chessboard used.
Figure 1: 6 DOF robot manipulator with controller, and the chessboard used.
Figure 2: Brief description of the architecture of the autonomous chess-playing system.
Figure 2: Brief description of the architecture of the autonomous chess-playing system.
Figure 3: Sampling process for the color detection of the pieces.
Figure 3: Sampling process for the color detection of the pieces.
Figure 4: Training set used for the decision tree algo- rithm. White pieces (red), black pieces (green) and empty squares (blue).
Figure 4: Training set used for the decision tree algo- rithm. White pieces (red), black pieces (green) and empty squares (blue).
Figure 5: Left, robot gripper and camera. Right, LED support mounted around the camera.
Figure 5: Left, robot gripper and camera. Right, LED support mounted around the camera.
Figure 6: Flowchart of the illuminator design and printing process.
Figure 6: Flowchart of the illuminator design and printing process.
Figure 7: Left, setup used to estimate the LED pa- rameters. Right, values measured in each point of the grid.
Figure 7: Left, setup used to estimate the LED pa- rameters. Right, values measured in each point of the grid.
Figure 9: Left, illumination level on the chessboard. Right, orientation of each diode. Once a good enough distribution is obtained, the coordinates of the location of the LEDs in the sup- port and the orientation vector for each LED are ex- ported into a text file, which is in turn imported into an OpenSCAD script that is responsible for generat- ing the 3D model of the support of the diodes, generat- ing three-dimensional holes in a user-specified shape where each LED will be mounted with the optimal ori- entation. The shape can be indicated by the user, as shown in Fig. 10, where different support shapes are
Figure 9: Left, illumination level on the chessboard. Right, orientation of each diode. Once a good enough distribution is obtained, the coordinates of the location of the LEDs in the sup- port and the orientation vector for each LED are ex- ported into a text file, which is in turn imported into an OpenSCAD script that is responsible for generat- ing the 3D model of the support of the diodes, generat- ing three-dimensional holes in a user-specified shape where each LED will be mounted with the optimal ori- entation. The shape can be indicated by the user, as shown in Fig. 10, where different support shapes are
Figure 10: Some of the LED supports printed and tested.
Figure 10: Some of the LED supports printed and tested.
Calculated board state (posterior to the movement)
Calculated board state (posterior to the movement)
Figure 11: Difference matrix for a basic movement.
Figure 11: Difference matrix for a basic movement.
Figure 12: Algorithm execution time comparison.
Figure 12: Algorithm execution time comparison.
Table 1: Classification matrices for the implemented algorithms. to attend other tasks simultaneously. The second rea- son may be related to a greater number of accesses to memory, explained naturally by the greater amount of data used by this algorithm.
Table 1: Classification matrices for the implemented algorithms. to attend other tasks simultaneously. The second rea- son may be related to a greater number of accesses to memory, explained naturally by the greater amount of data used by this algorithm.
Table 2: Results using LED illuminator Table 3: Results not using LED illuminator
Table 2: Results using LED illuminator Table 3: Results not using LED illuminator
Related topics:
Computer ScienceArtificial IntelligenceComputer VisionDecision MakingHuman-Robot Cooperation
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved