RaspberryPi Camera Remote Kivy User Interface
Python
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
README.md
android.txt
main.py Cannot use str.strip on unicode in kivy 1.8 - use object strip instead. Feb 6, 2014
piKamCommon.py Extracted the common stuff. Jan 12, 2014
piKamPicamServer.py Tidy ups and minor performance improvements. Jan 21, 2014
piKamServer.py Tidy ups and minor performance improvements. Jan 21, 2014
piKamSplash.jpg Add a splash image (mainly for testing that images display). Oct 13, 2013
pikam.ini
pikam.kv
screenshot.jpg Smaller - but more detailed Sep 22, 2013
screenshot2.jpg
screenshot3.jpg Tidy up rotation code until Kivy Android rotation can be automatic. Jan 19, 2014

README.md

piKam - Raspberry Pi Kivy User Interface

Recent changes

2014/02/06

  • Fix crash at startup due to UFT-8/string incompatibility for Kivy 1.8 (Android Kivy Launcher).

2014/01/20

  • The user interface app can now be run on the Raspberry Pi without running the server (limited by Kivy bugs - see Issues).
  • The app now has horizontal and vertical layouts (configuration switch).

2014/01/12

  • Now with live/zombie preview.
  • Eliminated to need to write thumbnails and previews to disk.
  • Adjusted +-ev values (the docs for raspistill seem wrong on value range).
  • New piKamCommon.py to install on the client and server.

Introduction

The piKam python Kivy application provides a remote interface to a Raspberry Pi camera. You can run piKam on any supported Kivy platform (Linux, Android, Windows, etc) and remotely drive the camera on a Raspberry Pi.

      

Warnings/Issues:

  • No security, no encryption (if your network has other users, they can also connect to the server).
  • The picam-based server's zoom isn't fully working.
  • The live view is semi-live, it's slow, perhaps a zombie-view more than a live-view.
  • Kivy settings don't seem to support placing restrictions on values and I haven't yet coded any sanity checks of my own.
  • Switching to scene mode Auto seems to improve brightness/contrast.
  • Negative ev sometimes doesn't seem to do much.
  • Running multiple raspistills at the same time can cause the camera to lock up (it's best not to run more than one client at a time).
  • Although the application can be run on the Raspberry Pi, due to bugs in the port of Kivy to the Raspberry Pi its not as usable the other platforms.
  • Until Kivy supports Android rotation events, the App does not auto-rotate.

Prerequisites

On your RaspberryPi - assuming Raspbian

Install the python-twisted package (used for client/server networking).

sudo apt-get install python-twisted

PiKam inclues a server that uses the raspistill command to drive the camera. If you setup your Raspberry Pi so that raspistill works, then piKam shold work too.

Optional picam support

I've included an alternate server that directly accesses the camera via Sean Ashton's picam module. I'd recommend using this module if you're interested in a server that is a bit more responsive.

If you wish to use the picam module then follow Sean's installation instructions at https://github.com/ashtons/picam/blob/master/README.md

Picam requires the python image library, which can be installed by running the following command:

   sudo apt-get install python-imaging-tk

On your selected Kivy client platform

I use piKam on Linux and Android, but it would likely run on any platform supported by Kivy.

Linux Kivy Client:

Install Kivy - most distributions seem to have a kivy package available. On OpenSUSE this seemed to install everything needed including Twisted.

Android Kivy Client:

Install the "Kivy Launcher" app from the Play Store.

Raspberry Pi Kivy Client:

Getting kivy running on a Raspberry Pi is still a bit tricky:

Install piKam and Test

Having installed the prerequisites, you can now install PiKam...

On your Raspberry Pi

Copy piKamServer.py and piKamCommon.py to your Raspberry Pi. Start the server - at the shell prompt:

     python piKamServer.py

If you wish to use the picam python module for direct camera access, then also copy piKamPicamServer.py to your Raspberry Pi and run that instead:

     python piKamPicamServer.py

The picam version is faster and probably to be preferred.

On your selected Kivy client platform

Linux PiKam Client:
  1. Create a folder called piKam (or anything you prefer). Copy all the piKam files to this folder.
  2. In a shell, cd to the piKam folder and enter:

    python main.py
    

    A GUI should fire up.

  3. Press F1 to bring up the piKam settings dialog.

  4. Set the address of you RaspberryPi and close settings.
  5. See if you can connect and "Capture" a photo.
Android PiKam Client:
  1. Find where "Kivy Launcher" put it's apps folder.
    • Lets say you find that Kivy Launcher is using /sdcard0/kivy/
    • Create a sub-folder /sdcard0/kivy/pikam
  2. Copy all the piKam files into this folder (skip piKamServer if you wish).
  3. Start the Kivy Launcher, it should list a piKam app, tap on it. A GUI should fire up.
  4. Tap the standard Android settings button/icon to bring up the piKam settings dialog.
  5. Set the address of you RaspberryPi and close settings.
  6. See if you can connect and take a photo.
Raspberry Pi Pikam Client

The PiKam client user interface can be run directly on the Raspberry Pi, but due to bugs in the port of Kivy to the Raspberry Pi you probably need a monitor with touch-screen support for it to be of much use:

  • Although the mouse works, the mouse cursor is not visible. This makes doing almost anything quite difficult - a touch screen would work around this issue because the pointer location would always be under your finger. As a temporary measure I've added code that plots the mouse location where it was last clicked.
  • It's better to run the GUI app in a console because terminating it via Control-C is more reliable in a console environment.
  • The keyboard doesn't seem to work inside Kivy, if you need to edit any settings stored in pikam.ini it can be manually edited.
  • Python, the Python Image Library, and Kivy together are rather draining on the Raspberry Pi. Preview refresh rates of less than 3 seconds are likely to reduce the interactivity of the GUI.
  • The GUI may be more responsive if it's configured to talk to a localhost piKamPicamServer running in a separate process - this reduces the load on the interpreter running the GUI and avoids any slowdowns due to the python global-interpreter-lock (edit the pikam.ini file and set the hostname).

If you have Kivy installed:

  1. Create a folder called piKam (or anything you prefer). Copy all the piKam files to this folder.
  2. Running a server is option, if you want PiKam to work directly without a server edit the pikam.ini file and make sure the server hostname is blank.
  3. In a shell, cd to the piKam folder and enter:

    python main.py
    

    A GUI should fire up. If you're running directly without a server, a preview should be visible.

  4. Mousing and clicking around may work, but you won't be able to see the mouse cusor. Control-C (possibly multiple times) will exit the GUI (hopefully).

(Note that you can't run a Raspberry Pi Kivy client remotely over the X11 protocal, so you can't work around the problems that way.)

Other clients

Kivy runs on a wide variety of platforms, installation would probably be similar to Linux.

Controls:

Mostly obvious.

  • Swipe left right on the image to goto prev/next on the image carousel. The carousel is limited to the 10 most recent images.
  • Photos are written to the client install folder.
  • Take a look at the settings menu for options/customisations (F1 on Linux, Tap Settings on Android).
  • If live view is enabled, the rightmost image on the carousel will display a low quality preview with the camera settings applied. If you move the view away from the rightmost image, the the live view network traffic suspends until the rightmost image is again visible.
  • The way live view is coded may mean you don't get much/any time to review a picture, but you can always scroll the carousel to suspend live view and review past images.
  • On Linux use F1 to get the settings, on Android use your settings key/symbol.