影像處理應用之簡介

影像處理應用之簡介

● 陳中亮

● 2017/12/12

簡介

What is Digital image processing ?

Digital image processing(數位影像處理) is the use of

computeralgorithmsto performimage processingondigital images.

As a subcategory or field ofdigital signal processing, digital image processing has many advantages overanalog image processing. It allows a much wider range of algorithms to be applied to the input data and can avoid problems such as the build-up of noise and signal distortion during processing. Since images are defined over two dimensions (perhaps more) digital image processing may be modeled in the form ofmultidimensional systems.

-- https://en.wikipedia.org/wiki/Digital_image_processing

Examples (Lena)

平滑化

影像幾何變換 伸縮、旋轉

影像幾何變換 伸縮、旋轉

Examples

Dilation(膨脹) ↔ Erosion(侵蝕)

Examples

邊緣偵測

其他(模糊、邊緣偵測…)

Examples

Hough line transform 霍夫線變換

透視變換(Perspective Transformation)

Examples

人臉偵測

Basic Concepts

影像即矩陣 0~255, 0 or 1

Basic Concepts

A colour image has 3 2-d matrices for R, G , B

Basic concepts

Example (0黑 ~ 255白)

Basic operations

 Reading and displaying images

 Writing images

Core module

 個別像素

 對比、亮度調整

g(x) = a f(x) + b

a = Contrast parameter b = Brightness parameter

Core module

Contrast and Brightness example

a =2.2 b=50

Core module

Drawing functions

Lines Circles Ellipses Polygon

Image Processing

平滑化(Smoothing) – Removes noise filters : Gaussian, Median, Bilateral,...

- medianBlur - GaussianBlur - bilateralFilter

Image Processing

Smoothing

Image Processing

Erosion and Dilation 用於減弱或突顯特徵

Erode + Dilate = Removal of stray marks

Erosion

Dilation

Image Processing

直方圖

Image Processing

直方圖均勻化 – 改進對比

Image Processing

邊緣偵測

Image Processing

Sobel Edge Detector

Image Processing

Laplace Edge Detector

Image Processing

Canny Edge Detector

Best edge detector available

Uses more advanced intensity gradient based methods

特徵偵測(Feature Detection)

The following 3 are considered to be keypoints in an image

1) Edges 邊緣

2) Corner 角點(also known as interest points) 3) Blobs (also known as regions of interest )

Once the features have been found, these features are

「described」.

Feature Detection

Feature Extraction: SURF, SIFT, BRIEF

Feature Descriptors: SURF, SIFT, BRIEF, STAR Matchers: FLANN, BruteForce

Template Matching

(圖樣比對,樣板比對,模板匹配)

Foreground Extraction(前景提取)

Corner Detection(角點偵測, 角點檢測)

Feature Match 特徵匹配

How we teach computers to understand pictures

李飛飛

https://www.youtube.com/watch?v=40riCqvRoMs

Image Processing v.s. Computer Vision

40

Image Processing 影像處理

Computer Vision 電腦視覺

Low Level

High Level

Acquisition, representation, compression, transmission image enhancement edge/feature extraction Pattern matching image "understanding“

(Recognition, 3D)

41

42