actfw 1.0.1

Actcast Application Framework

Actcast Application Framework for Python

This package provides a Python API for developing Actcast apps.

Installation

for Raspberry Pi

sudo apt-get update
sudo apt-get install -y python3-pil python3-setuptools python3-wheel picamera
pip3 install .

Usage

Construct your application with a task parallel model

• Application
  • actfw.Application : Main application
• Workers
  • actfw.task.Producer : Task generator
    • actfw.PiCameraCapture : Task generator which generate capture image as task
  • actfw.task.Pipe : Task to Task converter
  • actfw.task.Consumer : Task terminator

Each worker is executed in parallel.

User should

• Define subclass of Producer/Pipe/Consumer

class MyPipe(actfw.task.Pipe):
    def proc(self, i):
        ...

• Connect defined worker objects

p  = MyProducer()
f1 = MyPipe()
f2 = MyPipe()
c  = MyConsumer()
p.connect(f1)
f1.connect(f2)
f2.connect(c)

• Register to Application

app = actfw.Application()
app.register_task(p)
app.register_task(f1)
app.register_task(f2)
app.register_task(c)

• Execute application

app.run()

Please, see and try examples.

Example

• example/hello : The most simple application example
  • Use HDMI display as 640x480 area
  • Capture 320x240 RGB image
  • Draw "Hello, Actcast!" text
  • Display it as 640x480 image (with x2 scaling)
  • Notice message for each frame
  • Support application setting
  • Support application heartbeat
  • Support "Take Photo" command
• example/grayscale : Next level application example
  • Use HDMI display as 640x480 area
  • Capture 320x240 RGB image
  • Convert it to grayscale
  • Display it as 640x480 image (with x2 scaling)
  • Notice message for each frame
  • Support application setting
  • Support application heartbeat
  • Support "Take Photo" command
• example/parallel_grayscale : Paralell processing application example
  • Use HDMI display as 640x480 area
  • Capture 320x240 RGB image
  • Convert it to grayscale
    • There exists 2 converter task
    • Round-robin task scheduling
  • Display it as 640x480 image (with x2 scaling)
  • Notice message for each frame
    • Show which converter processes image
  • Support application setting
  • Support application heartbeat
  • Support "Take Photo" command