Félix Miguel Trespaderne

LWDOS: language for writing descriptors of outline shapes

A new method to obtain a rough description of a 3-D object from its outline shape is presented in this paper. Firstly the outline shape is split up into parts that are related in a connectivity graph. The boundary features of every part (lines, curves) and the information about its junction with other parts are extracted in order to provide a semantic content to the outline shape graph. A specific language to describe this graph has been developed. Finally, some results obtained applying our approach over real images are presented and discussed.

Visually guided robot for radiator sealing

A robotic vision-guided system that permits to manipulate different mixed parts is described in this paper. In order to work on a wide range of sizes and without stopping the products the industrial robot is equipped with two vision systems. A fixed vision system acquires a global image of the approaching radiators and computes a first pose estimation. This estimation is used for the robot to approach to the part and a second system, mounted on the robot arm, acquires a closer image to determine precisely the position and orientation just before the manipulation is performed. The developed system is working nowadays in an industrial line making it possible to robotize a complex task without using costly and inflexible fixtures.

Inspection of Stamped Sheet Metal Car Parts Using a Multiresolution Image Fusion Technique

This paper presents an image processing algorithm for on-line inspection of large sheet metal car parts. The automatic inspection of stamped sheet metal is not an easy task due to the high reflective nature of the material and the nearly imperceptible characteristics of the defects to be detected. In order to deal with the ubiquitous glints, four images of every zone are acquired illuminating from different directions. The image series is fused using a Haar transform into a single image where the spurious features originated by the glints are eliminated without discarding the salient information. Our results clearly suggest that the proposed fusion scheme offers a powerful way to obtain a clean image where these subtle defects can be detected reliably.

Parallel computing for real time gauze detection in laparoscopy images

Automatic systems are crucial for expanding the capabilities of surgeons, especially in minimally invasive surgery. In this paper we present an image processing system that detects in real time the surgical sponges in the video captured by the endoscope. This application is the first part of an automatic system to avoid the risk of inadvertent retention of gauzes in the patient body after the operation. The detection algorithm present real time capabilities using an affordable multicore hardware.

INSPECTION OF METALLIC STAMPED PARTS USING AN IMAGE FUSION TECHNIQUE

Abstract This paper presents a machine vision system for on-line inspection of large stamped car parts. The vision system includes a CCD camera and an illumination device that permit to obtain two images of the same zone illuminated from two different directions. The acquired images are fused using an energy function in order to eliminate the spurious features originated by glints and shadows of the metallic surface. The defects, if they are present, remain in the fused image and can easily be pointed out.

System for the production control and automatic packaging of plastic air sleeve guides

This paper describes the functioning of a manufacturing cell, in which several complex systems have been joined, such as: a robot, a computer vision system, and a lathe. To allow communication between these systems in an organised manner, diverse technologies have been used, varying from a field bus to the intranet of the company. The robot takes the role of master, coordinating the functioning of the rest of subsystems. The computer vision takes charge calculating the position and orientation of the pieces of injected plastic and communicating them to the robot. The robot gathers the pieces and takes them up to the lathe. Once the process of the lathe has finished, the robot gathers again the piece and deposits it in the packaging system. The cell is endowed with the possibility of remote adjustment and monitoring across the companys intranet. The cells functioning conditions can be modified from the laboratory, to assure that the pieces made are within the quality limits specified by the clients.

Applying vision guidance in robotic fresh meat handling

Abstract In this paper we present a visually guided robotic system to manipulate irregular pieces of meat in a cooking ham manufacturing process. Visual sensing and manipulation have been combined in an open-loop fashion in order to avoid specialized and expensive pipelined pixel processing hardware. Special attention has been paid to the image analysis introducing a robust contour algorithm. This algorithm is well suited for real time applications due to the inherent local processing of the image near the contour elements. Using this contour a trajectory that guarantees a secure apprehension of this slippery product is determined.

Hierarchical Stereo Matching Using Chromatic Information

Abstract The key issue in making stereo vision practical is to find a combination of algorithms and hardware that led to reliable, real-time range estimation with a computer system small enough to use on robotic vehicles. In this paper, a matching algorithm to solve the central problem of binocular stereo vision systems is proposed. The algorithm utilizes two primitives of different abstraction undera hierarchical scheme: regions and linear edge segments. The hierarchical scheme greatly improves the efficiency because provides only the needed depth information with the needed degree of resolution. Chromatic information has been used. Color significantly reduces the ambiguity between potential matches and increases the accuracy of the resulting matches.

Stereo camera calibration with distortion model and using forgetting factor

In this paper, we present a calibration technique for cameras with several types of distortion lenses. The algorithm decomposes the calibration parameters into nondistortion parameters and distortion parameters, and the estimation is performed using the weighted least-squares method. The validity and the performance of the calibration algorithm is tested on two stereo cameras mounted in a stereoscopic bank. Experimental results show the feasibility and the performance of the calibration algorithm.