[PYTHON] RaspberryPi Zero and OpenCV

This is also an article under investigation.

The definitive version of Raspberry Pi "Raspberry Pi Zero W" has finally passed the skill Also on sale at Switch Science, which is popular with manufacturers

$ 5 Raspberry Pi begins domestic handling for full-scale spread Raspberry Pi Zero, sold for 702 yen Now it's finally easier to get.

Motor controller expansion board for Raspberry Pi Zero "ZeroBorg" Is also coming out.

Installing OpenCV on your Raspberry Pi Zero http://www.pyimagesearch.com/2015/12/14/installing-opencv-on-your-raspberry-pi-zero/

According to this article I have OpenCV-Python installed. OpenCV is installed from the source code. I'm running CMake on the command line. make It takes almost 10 hours.

** $ 5 computer "Raspberry Pi Zero" initial setup and benchmark (1/3) **

http://monoist.atmarkit.co.jp/mn/articles/1512/07/news062.html

Introducing the Raspberry Pi Zero, an ultra-compact computer for only $ 5 The ultra-compact computer "Raspberry Pi Zero", whose price has been reduced to $ 5 (about 610 yen), has appeared. http://www.itmedia.co.jp/news/articles/1511/26/news117.html

Equipped with Broadcom BCM2835 with ARM11 core operating at 1GHz, 512MB memory, microSD memory card slot, mini-HDMI port, microUSB port, expansion terminal, etc. The size is 65 x 30 x 5 mm, which is the smallest for a Raspberry Pi.

There is a 40 Pin Unpopulated GPIO Header in the figure introduced.

** Foreign book Raspberry Pi Computer Vision Programming **

Is written assuming a normal Raspberry Pi. As you can see from the table of contents Computer Vision is introduced within the framework of OpenCV-Python. For those who have already mastered OpenCV-Python in other books etc., it will be less thankful. It may be recommended for those who are new to the OpenCV-Python framework and don't bother with English.

** Foreign book Learning Raspberry Pi Samarth Shah **

Or [Amazon link](http://www.amazon.co.jp/Learning-Raspberry-Pi-Samarth-Shah/dp/1783982829/ref=sr_1_1?s=english-books&ie=UTF8&qid=1451137049&sr=1- 1) Describes how to use GPIO and OpenCV-Python. Pages such as object detection (face detection / eye detection) and object tracking can also be seen in the table of contents. I would love to see how well these work with the resources of the Raspberry Pi Zero.

When image recognition technology is in place, the application doesn't necessarily require console input, nor does it have to show the results on a display. The completed configuration can be keyboardless or displayless. Of course, during development, there seems to be no inconvenience in using a USB hub, connecting a USB keyboard or mouse, and displaying the results on HDMI.

Usage I envision -Develop algorithms centered on OpenCV-Python. (Windows PC is also acceptable) ・ GPIO, USB-serial and other methods with RaspberryPi2 (one of the sales destinations) Develop applications with IO in. -After that, use the same GPIO and make it easy to move with RaspberryPi Zero. -Move with RaspberryPi Zero.

** Related information: OpenCV on ARM board **

Armadillo

How to: Image processing using OpenCV with Armadillo-810

Armadillo-800

Zynq Accelerating OpenCV functions with FPGA circuits is possible with FPGAs from both Altera and Xilinx. Speed up can be achieved by replacing the bottleneck OpenCV function with the OpenCV compatible FPGA design provided by both companies. With ARM-based CPUs and standardized bus designs available, it's a lot easier than when you were considering circuitization on an FPGA alone. The following resources show that you can use OpenCV compatible functions already created by FPGA vendors without knowing VerilogHDL at all.

Accelerate OpenCV applications in Zynq with the Vivado HLS Video Library http://japan.xilinx.com/support/documentation/application_notes/j_xapp1167.pdf

Using OpenCV and High-Level Synthesis in Vivado (Dubbed in Japanese) http://japan.xilinx.com/training/vivado/leveraging-opencv-and-high-level-synthesis-with-vivado.htm

Because the Raspberry Pi Zero is small, lightweight, consumes less power, and has high computing power. It seems to be compatible with Drone.

I found the following article.

How to build a $200 smart drone with the Pi Zero http://www.zdnet.com/article/how-to-build-a-200-smart-drone-with-the-pi-zero/

A Raspberry Pi Zero drone https://www.youtube.com/watch?v=LWwkB6hGD4M

Strapping on a Raspberry Pi Zero to a drone! http://www.recantha.co.uk/blog/?p=14978

Wireless can be dealt with by using GW-450D2 (wireless LAN dongle). Use GW-450D2 (wireless LAN dongle) with Raspberry Pi ZERO

Reference article OpenCV memorandum [Building OpenCV 3.1.0 + contrib 3.1.0 on Raspberry Pi 2] (http://iwaki2009.blogspot.jp/2015/12/raspberry-pi-2-opencv-310-contrib-310.html)

Raspberry Pi Zero (W) Beginning HW acquisition, SSH, browser environment

Run OpenCV on Raspberry Pi (Zero (W)) and save the image from USB Camera as JPEG