Pick and Place is an important part of industrial processing. For example, it allows the assembly of elements by a robot. These systems working with cameras and a robot must be adaptable to the type of objects to identify as well as easy to use. An operator is not expected to have the skills to set and to use pick and place machines. In this paper, we propose an easily usable algorithm called HSP2, Harmony Search for Pick and Place, to simplify bin-picking by using an evolutionary algorithm to set up cameras and identify the location of thin objects inside a scene so as to send their coordinates to the robot. The algorithm has been tested on camera calibration and the detection of thin objects.