GE 113 REMOTE SENSING Topic 7. Image Enhancement - WordPress

GE 113 – REMOTE SENSING Topic 7. Image Enhancement Lecturer: Engr. Jojene R. Santillan jrsantillan@carsu.edu.ph Division of Geodetic Engineering College of Engineering and Information Technology Caraga State University

Outline Part 1. Image Enhancement Concepts Part 2. Contrast Manipulation Techniques Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 2

Expected Outcomes The students would be able to: – Learn the concepts behind image enhancement – Identify the various computer-assisted procedures of image enhancement – Learn how to conduct the computer-assisted procedures through laboratory exercises Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 3

PART 1. IMAGE ENHANCEMENT CONCEPTS Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 4

Image Enhancement The goal is to improve the visual interpretability of an image by increasing the apparent distinction between the features in the scene. Why do we need a computer to do the enhancement? – Our eyes are poor at discriminating the slight radiometric or spectral differences that may characterize such features – With computers, these slight differences can be visually amplified to make them readily observable by our eyes. Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 5

Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 6

Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 7

Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 8

Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 9

Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 10

Types of Image Enhancement Operations Point Operations – Brightness values of each pixel in an image data are modified independently Local Operations Brightness values of each pixel in an image data are modified based on neighboring brightness values Note: Either form of enhancement can be performed on single-band images or on the individual components of multi-image composites. Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 11

When are image enhancement techniques applied? Normally applied to image data after the appropriate image rectification and restoration procedures have been performed. Noise removal very important to conduct prior to image enhancement – Image enhancement techniques may enhance “noise” if they are not removed the interpreter will end up analyzing enhanced noise! Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 12

Categories of Image Enhancement Techniques Contrast Manipulation Techniques – Discussed in detail in Part 2 Spatial Feature Manipulation Techniques – Used to emphasize or deemphasize image data of various spatial frequencies Spatial frequency refers to the roughness of the tonal variations occurring in an image – These are “local” operations pixel values in an original image are modified on the basis of the gray scale/brightness/DN values of neighboring pixels – Examples: Spatial filters Multi-image Manipulation Techniques – Enhancements involving multiple spectral bands of imagery – Examples: Spectral ratioing Principal and canonical components transformation Vegetation components transformation Intensity-hue-saturation color space transformation Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 13

PART 2. CONTRAST MANIPULATION TECHNIQUES Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 14

Contrast Manipulation Focused on manipulating the brightness values/DNs of an image data to reveal specific or new information or to enhance existing image information Commonly used contrast manipulation procedures: – Gray-level thresholding – Level slicing – Contrast stretching These are all “point” operations Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 15

Gray-level Thresholding A segmentation procedure An input image band is segmented into two classes: – One class for those pixels having values below a defined gray level (DN) – One class for those pixels above this value The result is a binary classification This binary classification can then be applied to a particular image band data to enable display of brightness variations in only a particular class Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 16

Example: NIR Band of Landsat 7 ETM Histogram of DN values of NIR Band DN Range: 0 – 30 water bodies Gray-scale Thresholded Image: Class 1: 0 -30 (Water) Class 2: 31 – 255 (Others) NIR Band of Landsat 7 ETM Showing only Class 1 (Water) Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT True Color Image Showing only Class 1 (Water) 17

Level Slicing An enhancement technique whereby the DNs distributed along the x axis of an image histogram are divided into a series of intervals or “slices”. All of the DNs falling within a ‘slice’ are then displayed at a single DN in the output image Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 18

NIR Band of Landsat 7 ETM Example: Histogram of DN values of NIR Band “Sliced” NIR Band of Landsat 7 ETM (6 classes) Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 19

Example: Sliced NIR Band (Water Portion only) Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 20

Example: Level slicing the TIR Band of Landsat 7 to show land surface temperature (LST) Image http://www.mdpi.com/2072-4292/7/4/4268/htm Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 21

Contrast Stretching (1) Recall: – An image can have DN values ranging from 0 to a maximum value depending on its radiometric resolution: E.g., an 8-bit image can have DNs ranging from 0 – 255 A 12-bit image can have DNs ranging from 0 – 4095 Etc. – When the image data are visualized on a screen of a computer, they are displayed as brightness values for each screen pixel A data pixel with a larger value is brighter than one with a smaller value However, unlike the image data, screen pixels can only have 256 unique brightness values (i.e., 0 to 255). This limitation prevents the data from being displayed with brightness exactly equal to their real (DN) value Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 22

Contrast Stretching (2) Stretching the image data refers to a method by which the data pixels are rescaled from their original values into a range that the monitor can display - namely, into integer values between 0 and 255. But what about contrast stretching? Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 23

Contrast Stretching (3) The parameters of the stretch can be adjusted to maximize the information content of the display for the features of interest this process is referred to as contrast stretching. Contrast stretching changes contrast in the image Contrast the relative differences in the brightness of the data values: – increasing an image's contrast means the dark pixels will become darker, and the bright pixels will become brighter – brightness difference between the two increases Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 24

Contrast Stretching as an Image Enhancement Procedure Used to expand the narrow range of brightness values typically present in an input image over a wide range of values Contrast stretching results to an output image or image display that is designed to emphasize the contrast between features of interest. Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 25

Types of Contrast Stretching (as implemented in various image processing software, e.g., Envi) Linear Linear 0-255 Linear 2% Gaussian Equalization Square root ALL OF THESE OPERATIONS RELY ON THE MANIPULATION OF THE IMAGE HISTOGRAMS Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 26

a graphical representation of the distribution of numerical data. To construct a histogram, the first step is to "bin" the range of values—that is, divide the entire range of values into a series of intervals—and then count how many values fall into each interval. The bins are usually specified as consecutive, nonoverlapping intervals of a variable. Number of Students What is a Histogram? Exam Score The bins (intervals) must be adjacent, and are usually equal size Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 27

What is a Image Histogram? A type of histogram that acts as a graphical representation of the tonal (“DN”) distribution in a digital image. It plots the number of pixels for each tonal/DN value. By looking at the histogram for a specific image a viewer will be able to judge the entire tonal distribution at a glance. Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 28

Linear Contrast Stretching Sets the image minimum and maximum DN values to values of 0 and 255, and stretches all other data values linearly between 0 to 255. Example: – If a band of an image has DN values ranging from 30 to 200, linear contrast stretching will expand the range such that when displayed/outputted to an image file, the new DN values will range from 0 to 255: Screen value of 0 will be assigned to 30 Screen value of 200 will be assigned to 255 All other values will be linearly stretched Algorithm: New DN DN’ [(DN – MIN) / (MAX – MIN) ] * 255 Where: DN original DN of a pixel MIN the image’s minimum DN value that will be assigned a new value of 0 MAX the image’s maximum DN value that will be assigned a new value of 255 Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 29

Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 30

Example: Linear Contrast Stretching Original Band 1 Stretched Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 31

Linear 0-255 Sets the image’s DN value of 0 to a new value of 0, and the image’s DN value of 255 to a new value of 255 “No stretching” Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 32

Example: Linear 0-255 Original Band 1 Stretched Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 33

Linear 2% Sets the highest and lowest 2% of the original image DN values to new values of 0 and 255, and it stretches all other data values linearly Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 34

Example: Linear 2% Original Band 1 Stretched Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 35

Gaussian Sets: – the original image’s mean DN value to a new value of 127, – the DN value 3 standard deviations below the mean value to a new value of 0, and – the DN value 3 standard deviations above the mean value to a new value of 255. Intermediate values are assigned new value using a Gaussian curve Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 36

Example: Gaussian Original Band 1 Stretched Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 37

Histogram Equalization Scales the original image DN values to equalize the number of DNs in each display histogram bin In this approach, image DN values are assigned to the display levels on the basis of their frequency of occurrence Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 38

Example: Histogram Equalization Original Band 1 Stretched Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 39

Square root takes the square of the input histogram and applies a linear stretch Original Band 1 Stretched Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 40

Questions or clarifications? Lecture Notes in GE 113: Remote Sensing TOPIC 7. IMAGE ENHANCEMENT 41