José Jerónimo M. Rodrigues

Phd in Computer Vision & Machine Learning, Carnegie Mellon | Portugal program

Main Menu

  • Home
  • CV/Resume
  • Bin-picking
  • ANSIG
  • Publications
  • Contacts

JoseRodrigues

José Jerónimo Moreira Rodrigues

PhD Student in ECE, within the Carnegie Mellon | Portugal program

Robotics Institute RI, Carnegie Mellon University CMU, Pittsburgh, USA, and 
Institute for Systems and Robótics (ISR), Instituto Superior Técnico (IST), Portugal

Email: jmoreira (@) cs.cmu.edu

Resume/CV

 


 

6D pose estimation of a textureless shiny object for bin-picking

bin image2Developed a method for 6D pose estimation of highly interlaced, shiny objects to automate bin-picking in assembly lines. The system fulfils the highly constrained requirements of assembly lines such as high speed and high accuracy, detecting parts with >99% of correctness, within 500ms. This is achieved by bringing together shape from shading concepts, large-scale machine learning and generalized hough transform. See more...

 

Professor Luís Vidigal Award

luis_vidigal

Award: Professor Luís Vidigal Award 2008 / 2009, best MSc Thesis within several courses in the Electrical Engineering and Computer Science scientific areas at IST.

Many thanks to professors Pedro Aguiar and João Xavier, advisors of the awarded thesis ANSIG — An analytic signature for permutation-invariant two-dimensional shape representation, José Rodrigues.


ANSIG

Developed a 2D shape representation (ANSIG) for automatic shape-based image classification.

ANSIG makes the comparison of arbitrary shapes trivial because it is invariant to the permutation of the landmarks.

ANSIG deals with translations, scale, rotations, and also with shapes represented by different number of points.

More info here. Software (matlab code) for automatic trademark retrieval and objec recognition is available.