File Converter Example

Basic Knowledge Requirements

Before diving into this example, here's a quick overview of what you'll need to be familiar with:

1. Python Programming: This example is based in Python, making use of various foundational constructs like functions, conditional statements, and object-oriented techniques.
2. OpenCV: A foundational understanding of the OpenCV library, particularly functions related to image decoding and display, as the example showcases how to visualize camera images from the log file.

The File Converter example will take the *.bin log file for a given Oak Camera Stream or snapshot captured by the Recorder app on an Amiga brain and convert it to either *.mp4 videos or *.jpg images for each view stream.

To successfully run this example, you must use your local PC, as the example won't work if executed directly from a brain (because of the popup window).

tip

You should typically expect 4 view streams per Oak Device ("rgb", "disparity", "left, "right").

1. Obtain a log file

Record your own

See the Recorder App Guide for instructions.

Download the log file

You can also download a pre-recorded log file to run this example with.

Click here to download

[Optional] Make a Data folder

We are going to make a folder that will store all of our log files, including the one you just downloaded.

cd <to-your-base-directory>
mkdir <data-dir>
cd <data-dir>
mv ~/Download/2023_09_29_17_52_35_070804_dubnium-durian.0000.bin

Now you should navigate (in your terminal) to the farm-ng-amiga repository.

2. Install the farm-ng Brain ADK package

3. Install the example's dependencies

Setup

cd farm-ng-amiga/

Recommended

Create a virtual environment

python3 -m venv venv
source venv/bin/activate

Install

cd py/examples/file_converter
pip install -r requirements.txt

4. Execute the Python script

Specify the log file, e.g.:

python main.py --file-name <path-to-file>/2023_09_29_17_52_35_070804_dubnium-durian.0000.bin

Additional args

Use the --help flag to see all possible arguments for using this tool.

# usage: Event file converter example. [-h] --file-name FILE_NAME
#                                      [--output-path OUTPUT_PATH] [--camera-name CAMERA_NAME]
#                                      [--view-name {rgb,left,right,disparity}] [--disparity-scale DISPARITY_SCALE]
#                                      [--video-to-jpg]
#
# optional arguments:
#   -h, --help            show this help message and exit
#   --file-name FILE_NAME
#                         Path to the `events.bin` file.
#   --output-path OUTPUT_PATH
#                         Path to the folder where converted data will be written.
#   --camera-name CAMERA_NAME
#                         The name of the camera to visualize. Default: oak0.
#   --view-name {rgb,left,right,disparity}
#                         The name of the camera view to visualize. Default: rbg.
#   --disparity-scale DISPARITY_SCALE
#                         Scale for amplifying disparity color mapping. Default: 1.
#   --video-to-jpg        Use this flag to convert video .bin files to a series of jpg images.
#                         Default is mp4.

For instance, you can change the camera that is played back from the default of oak0. E.g.,

python main.py --file-name events_09162022160753_000000.bin --camera-name oak1

You can convert video logs to a stream of jpg images, rather than the default mp4 videos.

python main.py --file-name events_09162022160753_000000.bin --camera-name oak1 --video-to-jpg

Or you can convert a "snapshot" log to one jpg per view.

python main.py --file-name cpy_data/farm_ng/2023_01_06_13_24_33_445932_snapshot_b42d218.bin --video-to-jpg

info

Video logs will not convert if the --snapshot flag is used. Similarly, snapshot logs will not convert if the --snapshot flag is not used.