Tribunal titular

Cargo

Departamento

Profesor

Presidente

ING.ELECTR.DE SIST. Y AUTOMATI

FOCES MORAN , JOSE MARIA

Secretario

ING.ELECTR.DE SIST. Y AUTOMATI

FUERTES MARTINEZ , JUAN JOSE

Vocal

ING.ELECTR.DE SIST. Y AUTOMATI

REGUERA ACEVEDO , PERFECTO

Tribunal suplente

Cargo

Departamento

Profesor

Presidente

ING.ELECTR.DE SIST. Y AUTOMATI

RIESCO PELAEZ , FELIX

Secretario

ING.ELECTR.DE SIST. Y AUTOMATI

ALAIZ RODRIGUEZ , ROCIO

Vocal

ING.ELECTR.DE SIST. Y AUTOMATI

GARCIA RODRIGUEZ , ISAIAS

Basic

Mohamed Elgendy. Deep Learning for Vision Systems. Manning. 2020

E.R. Davis. Computer Vision: Principles, Algorithms, Applications, Learning. Academic Press. 2017.
Francois Chollet. Deep Learning with Python. Manning. 2018 (2021 edition in preparation: https://www.manning.com/books/deep-learning-with-python-second- edition#toc)

Richard Szeliski. Computer Vision: Algorithms and Applications, Springer, 2011

(pdf available online: http://szeliski.org/Book/)

Pattern Classification (2nd Edition), by R.O. Duda, P.E. Hart, and D.G. Stork, Wiley-Interscience, 2000.

Alegre, E., Sánchez, L., Fernández, R.A. y Mostaza, J.C. (2003). Procesamiento Digital de Imagen: Fundamentos y Prácticas con Matlab. Secretariado de Publicaciones y Medios Audiovisuales de la Universidad de León.

González, R. C. y Woods, R. E. (2008). Digital Image Processing (Third Edition). Prentice Hall.

Complementary

Pajares, G y de la Cruz, J.M. (2001). Visión por Computador. Ra-ma.

González, J. (2000). Visión por Computador. Paraninfo.

González, R. C. y Woods, R. E. (1996). Tratamiento digital de imágenes. Addison-Wesley /Diaz de Santos.

Shapiro, L. & Stockman G. (2001). Computer Vision. Prentice-Hall.

Parker, J.R. (1997). Algorithms for image processing and computer vision. John Wiley & Sons, Inc.

Trucco, E. & Verri, A. (1998). Introductory Techniques for 3-D Computer Vision. Prentice-Hall.

Maravall, D. (1993). Reconocimiento de formas y visión artificial. Ra-ma

Davies, E.R. (1996). Machine Vision: Theory, Alforithms, Practicalities. Academic Press.

Code
A18131
B5614
B5619
B5623
B5625
B5627
C2	CMECES2 That students know how to apply their knowledge to their work or vocation in a professional manner and possess the skills that are usually demonstrated through the development and defense of arguments and the resolution of problems within their area of study.
C4	CMECES4 That students can transmit information, ideas, problems and solutions to both a specialised and non-specialised audience
C5	CMECES5 That students have developed those learning skills necessary to undertake further studies with a high degree of autonomy

	Competences
Understand the main concepts related to computer vision and digital image processing Understand and know how to program description techniques and object recognition based on machine vision Learning to use Python to program simple machine vision applications Be able to use computer vision techniques learned to solve new problems	A18131	B5614 B5619 B5623 B5625	C2 C5
	A18131	B5614 B5623	C4 C5
	A18131	B5614 B5619 B5623 B5625	C2 C4 C5
	A18131	B5614 B5619 B5623 B5625 B5627	C4 C5

Topic	Sub-topic
Block I. Lectures	Lesson 1: INTRODUCTION TO ARTIFICIAL VISION What is artificial vision. Its use Advantages and limitations. Lesson 2: BASIC CONCEPTS Image concepts, resolution, models and colour modes. Lesson 3: PREPROCESSING Main operations to pre-process images (arithmetic, logic, morphological). Lesson 4: EDGE DETECTION Edge concept. Edge detection methods. High pass filters for edge detection. Lesson 5: CORNER DETECTION Corner concept. Moravec detector. Harris Detector. Lesson 6: SEGMENTATION Segmentation concept. Use of filters as a form of segmentation. Thresholding segmentation methods. Region-oriented segmentation. Lesson 7: DESCRIPTORS Descriptor concept. Shape descriptors. Texture descriptors. Descriptors specified in local inventorial characteristics. BoVW. Image classification with BoVW. Lesson 8: MACHINE LEARNING Supervised learning. Main supervised learning methods. Clustering Evaluation methods. Lesson 9: DEEP LEARNING Fundamental parts of neural networks. Cost function. Neural network training. Optimization. Deep Learning for Image Classification.
Block II. Labs	Several practicals oriented to learn how to program in Python 3.X and, at the same time, to use this language to solve problems related to computer vision.

Methodologies :: Tests
	Class hours	Hours outside the classroom	Total hours
Problem solving, classroom exercises	8	12	20

Laboratory practicals	30	45	75

Lecture	18	27	45

	10	0	10

(*)The information in the planning table is for guidance only and does not take into account the heterogeneity of the students.

	Description
Problem solving, classroom exercises	Exercises done on the table, the blackboard and through slides and examples using Excel. Also, exercises could be left in Agora, with solutions, most of the times also with examples solved in Excel.
Laboratory practicals	The practicals are carried out using Python (3.X). Available in the laboratory F3 although it is recommended that each student also install it on her/his personal computer. A first lab about how to install and use of the environment will be conducted. Several labs aimed at learning basic Python combined with computer vision and machine learning concepts will be carried out later. Finally, some specific examples related with image classification will be the subject of the practicals. These examples will help students doing their final projects. Using Python 3.X from Anaconda installation environment (https://www.continuum.io/downloads). the Spyder will be used IDE included in the installation. The practices will be carried out following the scripts left in Agora. Doubts will be resolved by consulting the teacher. Your solution will be delivered in a task enabled for it in Agora.
Lecture	Theoretical sessions in the classroom using slides Presentations or documents corresponding to the materials of each lesson will be left in Agora and a forum for doubts will be set up. Some lessons will be accompanied by videos related to the concepts presented, some recorded by the teachers and others from internet resources that the teachers consider especially appropriate. Some lessons will be accompanied by a questionnaire with questions, which can be both theoretical and practical, whose deliveries will be evaluable. Datacamp platform courses, or similar ones, could be used to reinforce some of the lessons taught, with some courses being optional and others being mandatory.

	Description	Qualification
	55 points (out of 100) The subject will be evaluated by the implementation and presentation of a project that will be divided into the next four phases. 1. The student will perform a theoretical study of the subject, presenting a report that will deliver and will be evaluated. This report will contain a description of the problem, an explanation of its interest, and a review of published work. 2. As a result of the previous document, the student will select one of the reviewed implementations, or perform a self-implementation if he/she wishes, and evaluate the method with a known dataset or with a self-constructed dataset. With this implementation, the student will evaluate the method studied and will test, in a practical way, how it works. 3. The student will prepare and make a brief presentation with both theoretical and practical parts. The presentation is part of the evaluation and will be graded by the instructor. 4. After the presentation, a validation test will be conducted where each student will have to answer various questions related to the subject studied.	55
Others	30 points (out of 100): The tasks delivered by the students will be evaluated. The student must submit all the tasks indicated as mandatory that will be evaluated by the teachers. To pass the course, at least 50% of the mandatory tasks have to be correctly delivered. 15 points (out of 100): Participation in the course as a function of the intervention and student attitude in lectures, delivery of voluntary activities, help in the review of materials and their interaction with the teacher. The delivery of voluntary tasks will be specially taken into account.	45

Other comments and second call
Delayed deliveries will be subject of a penalty in the mark. All project deliverables will be further assessed by a validation test. At least 50% of the compulsory activities must be handed in and passed to pass the course. Students who do not pass the continuous assessment, could deliver what they lack in the period of the first ordinary call. Optionally, students could also pass the course in the second call, taking a theoretical and practical exam that will cover all the materials taught during the course. To pass in the second call, it will be compulsory to present the Course Project.