Samenvatting

This report is about implementing a pipeline in C++ in order to make it possible to perform online
hyperspectral analysis of tomatoes. The project is performed by a mechatronic student as a
graduation internship. The report will give insight in the design and realization of software which is
capable to record images as fast as possible with the use of concurrent programming. Going through
this report it will first describe the definition of the project with its requirements. After that it will
describe the preliminary design. Where with the use of diagrams the choices and the design of the
software and the hardware will be described. Once it is known how the setup will look like the
realization of the software and the hardware is described. At the end some test and the results will be
described and concluded.
In order to have a good understanding of the project the project definition will be described. First some
background information is given which explains what technologies are available and how these could
be used for the project. It explains the type of hyperspectral sensor that will be used and how the
hyperspectral data can contribute to the analysis of tomatoes. After this the problem is explained.
Once this is done the organization of the project is described. With the knowledge of the definition and
the problem of the project the requirements are defined.
Now that the problem and the project are defined and it is know what the requirements are the
preliminary design can be described. The preliminary design will help to get an understanding of the
configuration of the project. It will describe all the components that will be used and how they are
connected. First all the components of the hardware that are needed to make the setup will be
described. A conveyer belt is used to transport the tomatoes. A hyperspectral camera will be used to
record the images and a FPGA based board will be used to control the process. Block definition
diagrams will be used to describe the relation between the components. After this the components of
the software are described. There will be described how the Non-virtual interface will be used to
secure the maintainability and readability of the code. There will be explained how the pipeline works
and how it is constructed. It also explains how there will be secured that the pipeline will be executed
as fast as possible. Different diagrams like state machine diagrams describe the behavior of the
software.
Once there is clear how the setup will be constructed the design can be realized. There will be
described which products are used to realize the design of the hardware. After that there is described
how the different components of the software are implemented. There is explained what functions are
used and how they behave. The content of each specific task of the pipeline will be described. State
machine diagrams will be used to show the behavior of some tasks.
With the realized setup some test can be done. The most interesting characteristic of the setup is the
timing. Measurements of the time of the software are done with the use of the internal clock of the
processor. With the results of the measurements can be seen how long it will take to go through the
entire pipeline and how long it will take to perform the different tasks. At the end of the report the
project can be concluded with the use of the results and the realized parts. After this
recommendations are given about further steps that could be taken and which research would be
interesting to do.