《视觉计算》教学大纲

课程代码

045101901

课程名称

视觉计算

英文名称

Visual Computing（MOOC）

课程类别

选修课

课程性质

选修

学时

总学时：48 实验学时：16 实习学时：0 其他学时：0

学分

2.5

开课学期

第六学期

开课单位

计算机科学与工程学院

适用专业

计算机科学与技术（全英创新班）、（全英联合班）

授课语言

英文授课

先修课程

数学分析、图像处理

课程对毕业要求的支撑

本课程对学生达到如下毕业要求有如下贡献：

指标点1.1: 掌握数学、自然科学、工程基础和计算机专业知识，并能够用这些知识表述计算机工程问题，并建立具体对象的数学模型以及求解；

指标点4.1：能够基于科学原理，通过文献研究或相关方法，调研和分析计算机复杂工程问题的解决方案；

指标点2.1： 能够应用数学、自然科学和工程科学的基本原理，识别和判断计算机专业的复杂工程问题的关键环节，表述计算机专业的复杂工程问题；

课程目标

完成课程后，学生将具备以下能力：

（1）了解视觉信息和计算分析的基本原理，了解如何使用一些数学理论和方法进行图像和视频处理。[1]

（2）掌握频谱分析和数字滤波的基本知识，并有良好的基础进一步研究模式识别和计算机视觉。[1]

（3）了解视觉计算的基本概念，原理及有效的数学工具。[2]

（4）学生能够处理视觉计算中的一些基本问题。[2]

课程简介

视觉计算是计算机科学与技术学科的一门课程，它结合了频谱分析、数字滤波、图像处理、模式识别和计算机视觉。本课程对那些希望进一步学习和在计算机科学与技术开展初步研究的学生有很大帮助。学习本课程后，学生应该了解视觉信息和计算分析的基本原理，了解如何使用一些数学理论和方法进行图像和视频处理，掌握频谱分析和数字滤波的基本知识，并有良好的基础进一步研究模式识别和计算机视觉。教师使得学生了解视觉计算的基本概念，原理及有效的数学工具。学生能够处理视觉计算中的一些基本问题。

教学内容与学时分配

 1.理论教学部分

 a)背景

学时数：3学时

主要内容：对本课程的基本说明，包括数字图像处理、数字图像分析、计算机视觉以及相关的视觉计算专题。

重点：图像复原，图像增强，图像特征，傅里叶变换，谱分析与数字滤波。

难点：图像特征，傅里叶变换，谱分析与数字滤波。

要求：了解计算机视觉的基本概念及其各个领域的关系。


 b)视觉计算主要概念基础

学时数：6学时

主要内容：对计算机视觉的主要概念，任务，模型，和方法进行介绍。介绍人类视觉和计算机视觉的区分，常用的数据处理格式和工具。

重点：图像处理，图像识别模型，图像不变特征，词袋模型。

难点：图像退化模型。

要求：理解视觉计算中的基本概念。


 c)图像基础和基础图像操作

学时数：6学时

主要内容：介绍图像的矩形表达形式，基于矩阵的卷积运算，推导卷积与线性时不变系统间的关系。介绍图像的直方图表示，点运算，直方图的正规化，均衡化，直方图匹配，介绍图像噪声的类型和群运算。

重点：像素，调色板，分辨率，8-bit图，卷积，线性时不变，点运算，图像噪声，群运算。

难点：卷积，线性时不变，点运算和群运算。

要求：理解卷积操作，理解卷积与线性时不变之间的关系，掌握图像的基础操作。


 d)低维图像特征提取算子

学时数：6学时

主要内容：对介绍低维图像特征提取算子，包括一阶边缘检测子 Prewitt, Sobel, Canny算子，二阶边缘检测子LoG,DoG。

重点： Sobel, Canny算子， LoG和DoG算子。

难点：Canny算子，LoG和DoG算子。

要求：了解基础低微图像特征提取算子的原理。


 e)傅里叶变换

学时数：6学时

主要内容：介绍傅里叶级数，幅值和相位，从傅里叶级数推导出傅里叶变换，并推导傅里叶逆变换。举例计算函数的傅里叶变化。

重点：傅里叶级数，幅值和相位。

难点：傅里叶变换和傅里叶逆变换。

要求：理解傅里叶变换的原理。


 f)离散傅里叶变换

学时数：6学时

主要内容：介绍香农采样定理，离散傅里叶变换，离散傅里叶逆变换，并以信号压缩和信号去噪为例分析其中特点。介绍二维离散傅里叶变换和其在图像处理中的应用。分析幅值，相位的改变对图像的影响。。

重点：香农采样定理，离散傅里叶变换，离散傅里叶逆变换。

难点：离散傅里叶变换，离散傅里叶逆变换。

要求：理解离散傅里叶变换的原理。


 2.实验教学部分

 a)图像基础知识

学时数：4学时

主要内容：图像矩阵，色彩模式，比特图像，卷积。

要求：了解图像矩阵，色彩模式，比特图像，卷积的代码实现。


 b)图像处理

学时数：4学时

主要内容：图像噪声，图像去噪，图像直方图。

要求：了解图像噪声，图像去噪，图像直方图等图像处理相关概念和方法。


 c)傅里叶分析

学时数：4学时

主要内容：傅里叶级数，傅里叶变换

要求：使用MatLab实现傅里叶级数，傅里叶变换。


 d)离散傅里叶分析

学时数：4学时

主要内容：离散傅里叶变换，二维离散傅里叶变换

要求：使用MatLab实现离散傅里叶变换，二维离散傅里叶变换。

实验教学（包括上机学时、实验学时、实践学时）

有

教学方法

课程教学以课堂教学、实验教学、综合讨论、网络以及授课教师的科研项目与积累等共同实施。

考核方式

课程考核内容包括期末考试，实验及作业。期末考试占总成绩的70%，实验和作业的得分占总成绩的30%。

教材及参考书

现用教材：M.S. Nixon, A.S. Aguado.  Feature Extraction & Image Processing for Computer Vision. (3rd edition).

主要参考资料：

 [1] H. Ji. Mathematic in Visual Data Processing. National University of Singapore. 2011.

 [2] G. Aubert and P. Kornprobst, Mathematical problems in image processing,   Applied Mathematical Sciences, Vol 147, Springer-Verlag, 2006.

 [3] S. Mallat, A wavelet tour of signal processing, AP Professional, London,1997.

[4] I. Pitas, Digital image processing algorithm and Applications, John Wiley & Sons, New York, 2000.

制定人及制定时间

许勇，2019.4.1

 “Visual Computing（MOOC）” Syllabus

Course Code

045101901

Course Title

Visual Computing（MOOC）

Course Category

Elective Courses

Course Nature

Elective Course

Class Hours

48

Credits

2.5

Semester

6th

Institute

School of Computer Science and Engineering

 Program Oriented

Computer Science and Technology Full English Creative Class

Teaching Language

English

Prerequisites

mathematical analysis, image processing

 Student Outcomes

 (Special Training Ability)

 1. Research: with the ability to conduct investigations on the complex engineering problems based on scientific principles and adopting scientific methods, including the experiment designs, analyzing and interpretation of data, and to obtain valid conclusions by information synthesis.

2. Applying the Modern Tools: with the ability to develop, select and use the appropriate techniques, resources, and modern tools and IT tools, including prediction and simulation, to solve the complex engineering activities in information security and understand the limitations.

Course Objectives

 After teaching, students have the following abilities：

 (1)know the basic principle and analysis of visual information and its computing, understand how to use some mathematical theories and methods for image and video processing.[1]

 (2)know the basic knowledge of spectral analysis and digital filtering, and have a good base for further studying about pattern recognition and computer vision.[1]

 (3)make students to know the basic concepts of visual computing, and the principle and usage of effective mathematical tools.[2]

(4)Able to deal with some basic program in visual computing.[2]

Course Description

 Visual Computing is one of the courses of computer science and technology, which combines spectral analysis and digital filtering, image processing, pattern recognition and computer vision. This course is beneficial to those students for further study and start-up research in computer science and technology. After studying of this course, students should know the basic principle and analysis of visual information and its computing, understand how to use some mathematical theories and methods for image and video processing，know the basic knowledge of spectral analysis and digital filtering, and have a good base for further studying about pattern recognition and computer vision. Teacher should make students to know the basic concepts of visual computing, and the principle and usage of effective mathematical tools. The students are able to deal with some basic program in visual computing.


Teaching Content and Class Hours Distribution

 (1) Theory teaching

 a) Background

 Number of hours: 3 hours

 Main content: A basic description of the course, including digital image processing, digital image analysis, computer vision, and related math topics.

 Key points: image restoration, image enhancement, image features, image segmentation, Fourier transform, spectral analysis and digital filtering.

 Difficulties: image features, image segmentation, spectral analysis and digital filtering.

 Requirements: Understand the basic concepts of computer vision and its relationships in various fields.


 b) The main conceptual basis of visual computing

 Number of hours: 6 hours

 Main content: Introduce the main concepts, tasks, models, and methods of computer vision. Introduce the distinction between human vision and computer vision, commonly used data processing formats and tools.

 Key points: image processing, image recognition model, image invariant features, word bag model.

 Difficulties: Image degradation model.

 Requirements: Understand the basic concepts in visual computing.


 c) Image basic and basic image operations

 Number of hours: 6 hours

 Main content: Introduce the rectangular representation of images, matrix-based convolution operations, and derive the relationship between convolution and linear time-invariant systems. Introduce the histogram representation of the image, point calculation, normalization of the histogram, equalization, histogram matching, and introduction to the type of image noise and group operation.

 Key points: pixels, palette, resolution, 8-bit plot, convolution, linear time invariant, point arithmetic, image noise, group operations.

 Difficulties: convolution, linear time invariance, point operations and group operations.

 Requirements: Understand the convolution operation, understand the relationship between convolution and linear time invariance, and master the basic operations of the image.


 d) low-dimensional image feature extraction operator

 Number of hours: 6 hours

 Main content: Introduce low-dimensional image feature extraction operator, including first-order edge detector Prewitt, Sobel, Canny operator, second-order edge detection sub-LoG, DoG.

 Key points: Sobel, Canny operator, LoG and DoG operators.

 Difficulties: Canny operator, LoG and DoG operators.

 Requirements: Understand the principle of the basic low-micro image feature extraction operator.


 e) Fourier transform

 Number of hours: 6 hours

 Main content: Introduce the Fourier series, amplitude and phase, derive the Fourier transform from the Fourier series, and derive the inverse Fourier transform. An example is to calculate the Fourier variation of a function.

 Focus: Fourier series, amplitude and phase.

 Difficulties: Fourier transform and inverse Fourier transform.

 Requirements: Understand the principle of the Fourier transform.


 f) Discrete Fourier transform

 Number of hours: 6 hours

 Main contents: Introduce Shannon sampling theorem, discrete Fourier transform, discrete Fourier transform, and analyze the characteristics by signal compression and signal denoising. Introduce the two-dimensional discrete Fourier transform and its application in image processing. Analyze the amplitude, the effect of the phase change on the image. .

 Key points: Shannon sampling theorem, discrete Fourier transform, discrete Fourier transform.

 Difficulties: Discrete Fourier Transform, Inverse Discrete Fourier Transform.

 Requirements: Understand the principle of the discrete Fourier transform.


 (2) Experiment teaching

 a. Image Fundamentation

 Credit hours: 4

 Contents: Image matrix, Color Mode, Bit image, Convolution.

 Requirements: Learn how to handle Image matrix, Color Mode, Bit image and Convolution in MatLab.


 b. Image Processing

 Credit hours: 4

 Contents: Image noise, Image denoise, Image histogram.

 Requirements: Learn how to handle Image noise, Image denoise and Image histogram in MatLab.


 c. Fourier Analysis

 Credit hours: 4

 Contents: Fourier series, Fourier transform

 Requirements: Implement Fourier series and Fourier transform using MatLab.


 d. Discrete Fourier Analysis

 Credit hours: 4

 Contents: Discrete Fourier Transform,2D Discrete Fourier Transform

Requirements: Implement Discrete Fourier Transform and2D Discrete Fourier Transform using MatLab.

Experimental Teaching

Yes

Teaching Method

The course teaching is implemented in the classroom teaching, experimental teaching, comprehensive discussion, network teaching and the research projects and accumulation of the teacher.

Examination Method

The examination of this course includes final examination, experiments and homework. The final examination will take 70% of the total score, experiments and homework will take 30%.

Teaching Materials and Reference Books

 Textbook:

 M.S. Nixon, A.S. Aguado.  Feature Extraction & Image Processing for Computer Vision. (3rd edition).


 Reference Reading:

 1.H. Ji. Mathematic in Visual Data Processing. National University of Singapore. 2011.

 2.G. Aubert and P. Kornprobst, Mathematical problems in image processing,   Applied Mathematical Sciences, Vol 147, Springer-Verlag, 2006.

 3.  S. Mallat, A wavelet tour of signal processing, AP Professional, London, 1997.

4.  I. Pitas, Digital image processing algorithm and Applications, John Wiley & Sons, New York, 2000.

Prepared by Whom and When

Yong Xu, April 1st 2019


《视觉计算》实验教学大纲

课程代码

045101901

课程名称

视觉计算

英文名称

Visual Computing（MOOC）

课程类别

选修课

课程性质

选修

学时

总学时：48    实验：16     实习：0    其他：0

学分

2.5

开课学期

第六学期

开课单位

计算机科学与工程学院

适用专业

计算机科学与技术（全英创新班）、（全英联合班）

授课语言

英文授课

先修课程

数学分析、图像处理

毕业要求（专业培养能力）

本课程对学生达到如下毕业要求有如下贡献：

指标点1.1: 掌握数学、自然科学、工程基础和计算机专业知识，并能够用这些知识表述计算机工程问题，并建立具体对象的数学模型以及求解；

指标点4.1：能够基于科学原理，通过文献研究或相关方法，调研和分析计算机复杂工程问题的解决方案；

指标点2.1： 能够应用数学、自然科学和工程科学的基本原理，识别和判断计算机专业的复杂工程问题的关键环节，表述计算机专业的复杂工程问题；

课程培养学生的能力（教学目标）

完成课程后，学生将具备以下能力：

（1）了解视觉信息和计算分析的基本原理，了解如何使用一些数学理论和方法进行图像和视频处理。[1]

（2）掌握频谱分析和数字滤波的基本知识，并有良好的基础进一步研究模式识别和计算机视觉。[1]

（3）了解视觉计算的基本概念，原理及有效的数学工具。[2]

（4）学生能够处理视觉计算中的一些基本问题。[2]

课程简介

视觉计算是计算机科学与技术学科的一门课程，它结合了频谱分析、数字滤波、图像处理、模式识别和计算机视觉。本课程对那些希望进一步学习和在计算机科学与技术开展初步研究的学生有很大帮助。学习本课程后，学生应该了解视觉信息和计算分析的基本原理，了解如何使用一些数学理论和方法进行图像和视频处理，掌握频谱分析和数字滤波的基本知识，并有良好的基础进一步研究模式识别和计算机视觉。教师使得学生了解视觉计算的基本概念，原理及有效的数学工具。学生能够处理视觉计算中的一些基本问题。

主要仪器设备与软件

仪器设备：PC

软件： MATLAB

实验报告

实验报告需包含以下部分：

  1. 实验概况：需要简明扼要地介绍实验的目的、原理以及环境

  2. 实验过程：需要详细阐述实验的设计、过程以及结果，简述实验中遇到的问题以及解决方案

总结：总结实验的收获，包括对理论知识理解的加深、对研究方法的掌握和对现代工具使用等。

考核方式

本课程注重过程考核，成绩比例为：

实验报告：60%

平时表现：40%

教材、实验指导书及教学参考书目

现用教材：M.S. Nixon, A.S. Aguado.  Feature Extraction & Image Processing for Computer Vision. (3rd edition).

主要参考资料：

 [1] H. Ji. Mathematic in Visual Data Processing. National University of Singapore. 2011.

 [2] G. Aubert and P. Kornprobst, Mathematical problems in image processing,   Applied Mathematical Sciences, Vol 147, Springer-Verlag, 2006.

 [3] S. Mallat, A wavelet tour of signal processing, AP Professional, London,1997.

[4] I. Pitas, Digital image processing algorithm and Applications, John Wiley & Sons, New York, 2000.

制定人及发布时间

许勇，2019.4.1


《视觉计算》实验教学内容与学时分配

实验项目编号

实验项目名称

实验学时

实验内容提要

实验类型

实验要求

每组人数

主要仪器设备与软件

1

图像基础知识

4

主要内容：图像矩阵，色彩模式，比特图像，卷积。

要求：了解图像矩阵，色彩模式，比特图像，卷积的代码实现。

综合性

必做

1

仪器设备：PC

软件： MATLAB

2

图像处理

4

主要内容：图像噪声，图像去噪，图像直方图。

要求：了解图像噪声，图像去噪，图像直方图等图像处理相关概念和方法。

综合性

必做

1

3

傅里叶分析

4

主要内容：傅里叶级数，傅里叶变换

要求：使用MatLab实现傅里叶级数，傅里叶变换。

综合性

必做

1

4

离散傅里叶分析

4

主要内容：离散傅里叶变换，二维离散傅里叶变换

要求：使用MatLab实现离散傅里叶变换，二维离散傅里叶变换

综合性

必做

1


 “Visual Computing（MOOC）” Syllabus

Course Code

045101901

Course Title

Visual Computing（MOOC）

Course Category

Elective Courses

Course Nature

Elective Course

Class Hours

48

Credits

2.5

Semester

6th

Institute

School of Computer Science and Engineering

Program Oriented

 Computer Science and Technology Full English Creative Class

Teaching Language

English

Prerequisites

mathematical analysis, image processing

Student Outcomes (Special Training Ability)

 1. Research: with the ability to conduct investigations on the complex engineering problems based on scientific principles and adopting scientific methods, including the experiment designs, analyzing and interpretation of data, and to obtain valid conclusions by information synthesis.

2. Applying the Modern Tools: with the ability to develop, select and use the appropriate techniques, resources, and modern tools and IT tools, including prediction and simulation, to solve the complex engineering activities in information security and understand the limitations.

Teaching Objectives

 After teaching, students have the following abilities：

 (1)know the basic principle and analysis of visual information and its computing, understand how to use some mathematical theories and methods for image and video processing.[1]

 (2)know the basic knowledge of spectral analysis and digital filtering, and have a good base for further studying about pattern recognition and computer vision.[1]

 (3)make students to know the basic concepts of visual computing, and the principle and usage of effective mathematical tools.[2]

(4)Able to deal with some basic program in visual computing.[2]

Course Description

 Visual Computing is one of the courses of computer science and technology, which combines spectral analysis and digital filtering, image processing, pattern recognition and computer vision. This course is beneficial to those students for further study and start-up research in computer science and technology. After studying of this course, students should know the basic principle and analysis of visual information and its computing, understand how to use some mathematical theories and methods for image and video processing，know the basic knowledge of spectral analysis and digital filtering, and have a good base for further studying about pattern recognition and computer vision. Teacher should make students to know the basic concepts of visual computing, and the principle and usage of effective mathematical tools. The students are able to deal with some basic program in visual computing.


Instruments and Equipments

Equipment: PC

Software: MATLAB

Experiment Report

 The following parts should be included in the experiment report:

 1. Introduction: the purpose, principle and environment of the experiment should be concisely introduced.

 2. Procedure: the design, process and result of the experiment should be explained in detail. The problems encountered in the experiment and solutions are briefly described

3. Summary: summarize the results of the experiment, including the deepening of understanding of theoretical knowledge, the grasp of research methods and the use of modern tools.

Assessment

 Report：60%

Performance in lab：40%

Teaching Materials and Reference Books

 Textbook:

 M.S. Nixon, A.S. Aguado.  Feature Extraction & Image Processing for Computer Vision. (3rd edition).


 Reference Reading :

 [1] H. Ji. Mathematic in Visual Data Processing. National University of Singapore. 2011.

 [2] G. Aubert and P. Kornprobst, Mathematical problems in image processing, Applied Mathematical Sciences, Vol 147, Springer-Verlag, 2006.

 [3] S. Mallat, A wavelet tour of signal processing, AP Professional, London,1997.

[4] I. Pitas, Digital image processing algorithm and Applications, John Wiley & Sons, New York, 2000.

Prepared by Whom and When

Yong Xu, 2019.04.01

 “Visual Computing” Experimental Teaching Arrangements

No.

Experiment Item

Class Hours

Content Summary

Category

Requirements

Number of Students Each Group

Instruments, Equipments and Software

1

Image Fundamental

4

Contents:Image matrix, Color Mode, Bit image,Convolution.

Requirements: Learn how to handle Image matrix, Color Mode, Bit image and Convolution in MatLab.

Comprehensive

Elective

1

Equipment: PC

Software: MATLAB

2

Image Processing

4

Contents: Image noise, Image denoise, Image histogram

Requirements: : Get familiar with basic image processing concepts and methods.

Comprehensive

Elective

1

3

Fourier Analysis

4

Contents: Fourier series, Fourier transform

Requirements: Implement Fourier series and Fourier transform using MatLab

Comprehensive

Elective

1

4

Discrete Fourier Analysis

4

Contents: Discrete Fourier Transform,

2D Discrete Fourier Transform

Requirements: Implement Discrete Fourier Transform and

2D Discrete Fourier Transform using MatLab.

Comprehensive

Elective

1