miguel

About Computer Vision by Tecnalia

This author has not yet filled in any details.
So far Computer Vision by Tecnalia has created 79 blog entries.
31 Oct 2019

CONSTELLATION LOSS

2019-12-10T19:05:24+02:00Categories: Deep Learning|Tags: , , , |0 Comments

Deep learning diagnostic algorithms are proving comparable results with human experts in a wide variety of tasks, they still require a huge amount of well annotated data for training which is often non-affordable. Metric learning techniques have allowed a reduction on the required annotated data allowing few-shot learning. Existing deep metric learning loss functions have made possible generating models capable of tackling complex scenarios with the presence of many classes and scarcity on the number of images per class, not only work for classification tasks, but to many other clinical applications where measuring similarity is the key. Currently used state-of-the-art loss functions still suffer from slow convergence due to the selection of effective training samples that has been partially solved by the multi-class N-pair loss by simultaneously adding additional samples from the different classes. The constellation loss goes one step further by simultaneously learning distances [...]

25 Oct 2019

SCHOOL OF AI (BILBAO)

2019-12-10T19:05:31+02:00Categories: Computer Vision|Tags: , , |0 Comments

School of AI invited Tecnalia researchers Alfonso Medela and Miguel G. San Emeterio to give a talk on “Deep Learning”. School of AI is a non-profit organization whose goal is to provide on AI all over the world. At the moment they have settled in 400+ cities distributed around 80+ countries which are managed by a group of 800+ people around the world. The newly created Bilbao branch of School of AI invited our organization's researchers to its third event held in Bilbao to share some of their knowledge on Deep Learning. The took place at the facilities of the University of Mondragon at Bilbao. Alfonso and Miguel talked about “What is Deep Learning?”, Artificial Intelligence, Neural Networks, Machine Learning and the impact, current issues and future research lines.

28 Aug 2019

COURSE IN MEXICO 2019

2019-10-25T13:13:39+02:00Categories: Computer Vision|Tags: , , , , |0 Comments

Continuing the work done in 2018 in Nuevo Laredo, Tecnalia researchers have participated in training tasks in several courses offered at the Technological Institute of Saltillo, Mexico. Saltillo is the capital city of the State of Coahuila, located in northern Mexico where the main economic activities are generated by the manufacturing industry, services sector and transport; contributing 3% of national GDP. The government of the State of Coahuila hired Tecnalia to be part of the HUB Coahuila Innovation 4.0, a joint initiative by the Ministry of Economy of the State, the National Chamber of the Transformation Industry Delegation Coahuila Southeast (CANACINTRA) and the Technological Institute Saltillo (ITS). It was created to support the Industry 4.0 revolution in the State of Coahuila, and to be a dynamic agent of the industry through the transfer of knowledge between companies and academia. Within the first phase of this initiative, [...]

20 Mar 2018

An artifical retina

2019-10-25T13:19:57+02:00Categories: Computer Vision|Comments Off on An artifical retina

The human visual system is the model that has been used to develop an artificial system that mimics the human retina. A system that imitates the processing that takes place in the human eye has been developed based on the modelling of the neurons in the human retina. Within the natural evolution process, the human visual system has specialised in generating multiple information channels from the image captured by the photoreceptors. This way, it analyses moving elements as well as the shapes and the colours of objects that are in its field of vision. This system is especially relevant in the evaluation of the colours present in an image. The model provides better results than standard CIECAM02 on Munsell’ s samples (perceptually uniform space). These developments will help create more advanced and robust segmentation algorithms for their use in numerous applications in machine vision, thus being able to link [...]

20 Mar 2018

Georadar analysis of buried assets

2019-12-04T16:50:45+02:00Categories: Computer Vision|Comments Off on Georadar analysis of buried assets

Georadar prospecting is a pioneering technology in the location of buried structures. At present, most municipalities do not have an updated cartography or a precise location of piping and cabling. This creates a lot of problems when having to intervene, such as risks of gas pipeline explosions, damage to existing infrastructures and prolonged interruption of services. Tecnalia’s Computer Vision Group developed for Wideworld Geographic Services an automated tool that is capable of detecting and identifying buried assets based on the information provided by a georadar. The MUSA software manages the synchronized capture of the georadar signal with its movement and employs image enhancement algorithms for the captured images while at the same time automatically detects and characterizes buried assets based on advanced artificial intelligence algorithms. It also provides a precise location of the asset based on the information that is integrated in its GPS and GIS. To a large [...]

20 Mar 2018

Dimensional control of automotive tubes

2019-07-10T14:46:49+02:00Categories: Quality Control|Comments Off on Dimensional control of automotive tubes

Pipe manufacturing is a complex process where dimensional defects may occur due to several causes that may affect the product´s quality. A machine vision system can be employed to detect those defects before they reach the customer. Tecnalia´s Computer Vision Group has developed a complete machine vision system for Tubos Reunidos which is capable of measuring both sides of the tube after it has been introduced in its point of inspection. It can measure several dimensional parameters such as diameters, wall thickness, concentricity, eccentricity, to name a few, and can verify that these values are within the allowed range for each reference. The system is installed in the cold drawn factory located in LLodio and has been key in guaranteeing the quality of the tubes. This is a portable system which is housed in a sole enclosure which includes the lighting and capture systems as well as the mechanical [...]

20 Mar 2018

Intelligent wood classification

2019-07-10T14:46:10+02:00Categories: Quality Control|Comments Off on Intelligent wood classification

Wood is a living organism. Its quality control and assessment based on visual inspection is a very complex process. Tecnalia Computer Vision in collaboration with the Sustainable Construction Division has developed a system that provides solutions to this problem. The system has been developed to detect different types of defects and features that exist in wood, such as mineral stain, alburnum, stains, buds as well as defects caused by dirtiness in the plant. The system can also evaluate the visual appearance of the wood in accordance with the location and extent of the different problems. Timber is very variable and its assessment by traditional means is very difficult. The problem has been solved by means of algorithms based on perceptual models that not only characterise the different types of defects, but also estimate the perceptual quality of the wood based on the types of defects and its location within [...]

20 Mar 2018

Foam process control

2019-07-10T14:45:07+02:00Categories: Quality Control|Comments Off on Foam process control

The manufacturing of plastic foams is a two-day long process from the moment of the initial pour of the component substances in the mold until the foam block is obtained in its final shape. There are some verifiable parameters which can guarantee the product´s quality and the specific customer requirements, such as firmness, permeability or number of pores. However, these are only obtained at the end of the foaming process and can only be foreseen by an expert´s intuition. Therefore, it is important to be able to predict the final result from the first stage of the process as well as to be able to immediately modify, if necessary, the production variables and introduce corrections. Tecnalia´s Computer Vision Group, in collaboration with IPF Engineering, has developed for ICOA foam manufacturer a machine vision system installed in the first stage of the foam process. This system calculates the parameters that [...]

20 Mar 2018

Surfin´1.0 employed at Tubos Reunidos

2019-07-10T14:45:43+02:00Categories: Quality Control|Comments Off on Surfin´1.0 employed at Tubos Reunidos

Tecnalia developed for Tubos Reunidos a novel automated inspection application for surface defects in hot rolling/lamination processing based on a redundant special lighting system based on wavelengths that are sufficiently distant from the emission of the incandescent steel. This system inspects the surface by employing optic filters, safety systems and three high resolution line cameras that are housed in three protection casings. An image has been achieved of the development of the complete tube as if the latter were at room temperature. This system has been fully operational for the last three years and has created a product line that is being implemented in other important companies in the sector. Tubos Reunidos S.A. is a renowned seamless steel tube manufacturer whose products are mainly used in the oil & gas industry. The rolling mill/lamination process can sporadically produce surface marks and defects due to the wear & tear of [...]

20 Mar 2018

High resolution three dimensional control of tubes

2019-07-10T14:44:30+02:00Categories: Quality Control|Comments Off on High resolution three dimensional control of tubes

Tecnalia Computer Vision has developed for Tubos Reunidos a novel three dimensional control system that allows a three dimensional measurement of the bevelling in 100% of its production. This project developed a three dimensional rotating scanner (P-200931287) that was custom designed for the measurement of revolving elements. This scanner is composed of a rotary edge that turns a set of two lasers and a camera. This equipment obtains different sections of each tube. These sections allows reconstructing the three dimensional image where the dimensional control takes place. This system solves the problems associated with the automated analysis in the scanning of rotating elements by means of this new scanning technique which allows the automated reconstruction and processing. The controlled measures by Tubos Reunidos are: external diameter, internal diameter, thickness, height, depth and angle of both internal as well as external bevels, roundness, concentricity and eccentricity of the tubes with [...]

Load More Posts