Skip to content

avinavsrivastava/iiih-super-resolution

Repository files navigation

iiih-super-resolution

Monsoon 2017: PGSSP Team 11

Steps for Executing SRCNN.py

  1. Download SRCNN.py, train.zip, test.zip, and val.zip in local folder.
  2. Extract train.zip, test.zip and val.zip in train, test and val folders respectively.
  3. Use python 2.7.14
  4. Install all dependent packages Keras, scipy, numpy, os, timeit, cv2.
  5. Open terminal/command prompt and navigate to above download folder.
  6. Execute command "python SRCNN.py"
  7. Execution will take maximum 30 mins.
  8. test folder will contain inuput low resolution images, bicubic images and predicted images.
  9. Best PSNR achieved will be reported at the end of execution.

Steps for Executing SRKNN.py

  1. Download SRKNN.py, train.zip and test.zip, in local folder.
  2. Extract train.zip and test.zip in train, test folders respectively.
  3. Use python 3.6.1
  4. Install all dependent packages sklearn, scipy, numpy, os, matplotlib.
  5. Open terminal/command prompt and navigate to above download folder.
  6. Execute command "python SRKNN.py "
  7. Execution will take less than 5 mins.
  8. Input high resolution, reconstructed image and blurred input images are saved in the output folder path given in the command line prompt.
  9. PSNR will be reported on the console after reconstruction of every image in the test folder given in the command line prompt.

Steps for Executing SRSVR.py

  1. Download SRSVR.py and SVR.zip in local folder.
  2. Extract SVR.zip to get train , test and Result folders respectively.
  3. Use python 2.7
  4. Install all dependent packages PIL, skimage,sklearn, scipy, numpy, glob, matplotlib.
  5. Open terminal/command prompt and navigate to above download folder.
  6. Execute command "python SRSVR.py . The Train , Test and Result folders should be in the same path. K value is 10e4
  7. Execution will take less than 5 mins.
  8. Result folder will have the Low resolution Image and the corresponding reconstructed Image with LR and SR prefixed.
  9. PSNR ( Individual and Mean) will be reported on the console after reconstruction of every image as well as summary in command prompt

About

Monsoon 2017: PGSSP Team 11

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •  

Languages