ENVI Experimental Tutorial (4) Experiment 4, Remote Sensing Image Preprocessing - Fusion, Mosaic, Cropping

Experiment 4. Remote sensing image preprocessing - fusion, mosaic, cropping

1. Purpose of the experiment

1. Understand several methods of remote sensing image fusion.
2. Familiar with ENVI commonly used image fusion, mosaic and cropping tools
3. Master the main methods and operation steps of ENVI image fusion, mosaic and cropping.

2. Basic requirements of the experiment

1. Carefully read and master the procedure of this experiment.
2. Operation and application of fusion, mosaic and cutting modules on the computer.
3. Save and record the experimental results, and analyze and summarize.

3. Experiment time and place

1. Place:
2. time:

4. Experimental conditions

1. Hardware: PC computer (Windows 10 operating system)
2. Software: ENVI 5.1
3. Reference material: Chapter 4 of "ENVI Remote Sensing Image Processing Method"
4. Data used: ...\Chapter 4 Remote Sensing Image Preprocessing\ 5-8 Examples of image fusion, mosaic, cropping and fusion

5. Experimental content

1. Fusion of images from different sensors
2. Same sensor image fusion
3. Use the Seamless Mosaic image mosaic tool
4. ENVI rules and irregular graphics clipping

6. Matters needing attention

1. According to the actual needs of the project, determine whether integration is required, and select the corresponding integration method.
2. The general sequence of image preprocessing: radiometric calibration—>atmospheric correction—>geometric correction—>(fusion)—>mosaic/stitching—>cropping.

Seven, the main steps of the experiment

(1) Image Fusion
Different sensor image fusion

In Toolbox, open / Image Sharpening /Gram-Schmidt Pan Sharpening, select TM-30m.img as the low resolution image (Low Spatial) and bldr_sp.img respectively in the file selection box as the high resolution image (High Spatial), Click OK. Open the Pan Sharpening Parameters panel.

In the Pan Sharpening Parameters panel, select the sensor type (Sensor): Unknown, the resampling method (Resampling): Cubic Convolution, and the output format: ENVI.

(2), image mosaic
2.1 loading data

Click on the upper left of the Seamless Mosaic panel to add the image data to be mosaiced.


In the Data Ignore Value list, you can set the transparency value.
Check Show Preview in the upper right corner to preview the mosaic effect

2.2 Color correction
In the Color Correction option, check Histogram Matching

2.3 Seamlines and feathering
Select the drop-down menu Seamlines > Auto Generate Seamlines to draw seamlines automatically.
Pull down the menu Seamlines> Start editing seamlines to edit the seamlines.

2.4 Output results
(1) In the Export panel, set the resampling method Resampling method: Cubic Convolution;
(2) Set the background value Output background Value: 0; (
3) Select the output path of the mosaic result;
(4) Click Finish to execute the mosaic

(3) Image cropping
3.1 Rule image cropping
(1) File > Open to open the image Beijing_TM.dat, and display it stretched by Linear2%.
(2) File > Save As, enter the File Selection panel, select the Spatial Subset option, and open the cropping area selection function on the right.

Click Subset By File, click the Open file button, and select the vector data "vector.shp" as the clipping range

3.2 Irregular Image Cropping
3.2.1 Manually draw the clipping area
(1) Open the image Beijing_TM.dat, and display it stretched by Linear2%.
(2) Select the Beijing_TM.dat file in the Layer Manager, click the right mouse button, select New Region Of Interest, and open the Region of interest (ROI) Tool panel.
(3) Click the button in the Region of interest (ROI) Tool panel to draw a polygon on the image, and draw a polygon roughly within the scope of the Second Ring Road of the Old Imperial City of Beijing as the clipping area. You can modify the ROI Name, ROI Color, etc., or draw several polygons according to your needs. When drawing multiple ROIs, you can use it to delete.

(4) In the Region of interest (ROI) Tool panel, select File-> Save as, save the drawn polygonal ROI, and select the saved path and file name.
(5) In the Toolbox, open Regions of Interest/ Subset Data from ROIs.
(6) In the Select Input File dialog box, select Beijing_TM.dat to open the Subset Data from ROIs Parameters panel.

(7) In the Subset Data from ROIs Parameters panel, set the following parameters:
• Select Input ROIs: Select the vector file roi1 just generated
• Mask pixels output of ROI? : Yes
• Mask Background Value: 0
(8) Select the output path and file name, click OK to execute image cropping.

3.2.2 Clipping images from external vector data


8. Experimental results and discussion
Through fusion, we learned how to enhance the resolution of a multispectral image.
In the mosaic, the transparency and feathering radius of the image should be changed to make the mosaic of the image blend better.
By cutting, you can get the desired selection, and you can import vector files during the experiment to get the shape you want.