VP9-Video-Decoder

Download

git clone https://github.com/aisu-programming/VP9-Video-Decoder.git

Installation

npm i

Though I have checked but may have dependency issues.

Preprocess

Download the zipped data file under src/encoded_data.

Unzip the file and add an underscore (_) before the directory, the groups txt file, and the traversal json file.

Your files should look like this:

If the JSON files are not well formatted:

1. Edit the DATE variable to target date in src/1_format_jsondata.py.
2. Execute it: python 1_format_jsondata.py

Next, for groups txt file, change it to JSON format:

1. Edit the DATE variable to target date in src/2_convert_groups_txt.py.
2. Edit the import source and the DATE variable to target date in src/3_extract_groups_py.py.
3. Execute them: python 2_convert_groups_txt.py python 3_extract_groups_py.py

Usage

The main part of the codes are in src/App.tsx.

To change the target decoding date, simply edit the date at line 396.

To change the function between "Group" and "Traversal", edit the comment at line 397~400.

Run the Back-End App, then run the Front-End App:

1. python backend.py
2. npm start
3. You should see logs in both the console of the browser and the Back-End App.

Verification (Only for "Group" function)

To check that the total of decoded files is correct:

1. Edit the DATE variable to target date in src/4_check_jsondata.py.
2. Execute it: python 4_check_jsondata.py
3. It will create a JSON file named "verification.json" in the output directory "decoded_images/{DATE}".
4. Check that the number in the resulting dictionary in "verification.json" are same to the total in the output directory.