INSTALL.md

Installation

These are instructions to build CV Workbench from source. Below are requirements to set up the system to build and run. Once they are in place, the system may be built with:

$ cd src
$ make

This builds the main program cv-workbench and some additional utilities.

The software is assumed to run under Linux Ubuntu or compatible Linux version, such as Microsoft WSL2 Pengwin. The following Linux modules need to be installed.

gcc and g++

Install the gcc and g++ build utilities with:

$ sudo apt install build-essential

Json-c

Install the Json-c C/C++ Json parsing library under Linux with:

sudo apt install libjson-c-dev

or

sudo apt-get install libjson0 libjson0-dev

• It can be cloned and built from source (cv-workbench was developed using this).
• See the json-c documentation and reference.
• See the json-c tutorial and examples.
• See an additional tutorial with these examples.
• See JSON documentation at json.org.
• JSON can be formatted on a website, with the jq command-line app, or in some IDEs such as notepad++ or JetBrains Clion.
• Compile jcson-c C source files with the json-c/json.h include file and link with -ljson-c.

Berkeley DB

Install the Berkeley DB open source embedded database system from the zip source with these instructions.

• See alternate install instructions for the freesource version.
• You make need to install these packages:

sudo apt install autoconf libtool

• See the Oracle Berkeley DB documentation.
• Other documentation is at DZone
• See Berkeley DB tutorials and sample applications.
• Compile Berkeley C source files with the db.h include file and link with -ldb.

JPEG support

Install the Independent JPEG Group jpeg-9e library from source (also from github).

• See the source documentation.
• There is a wiki and more documentation.
• See the JPEG documentation.
• There is an example.c file in the source that is an example of using the library.
• Compile jpeg-9e C source files with the jpeglib.h include file and link with -ljpeg.

OpenCV

Install the OpenCV open source computer vision package to develop workbench applications with OpenCV.

• See the OpenCV documentation.
• Compile OpenCV C++ source files with the opencv2/opencv.hpp include file and link with various -lopencv_* libraries found in /usr/local/lib.

Sqlite3

Install the embedded relational database Sqlite3 under Linux with:

sudo apt install sqlite3

• It can be built from source.
• See these build instructions.
• A Windows ODBC driver can be installed for use by, e.g., Microsoft Access.
• See the sqlite3 tutorial.
• See the sqlite3 and C/C++ API documentation with a tutorial, an example, and source files.
• Compile sqlite3 C source files with the sqlite3.h include file and link with the sqlite3.o file generated from the above source with:

gcc -c sqlite3.c

• See these examples of building apps with sqlite3.

ESP32-CAM

To use the Arduino ESP32-CAM for image input, install the Arduino ESP32-CAM module in the Arduino IDE.

• See the ESP32-CAM camera webserver still photo capture project.