Prototyping profilometry on a microscopic level

#Prototyping     #Technology flexibiliy

GenPyCam was vital in developing a new machine to generate objects' depth maps accurately, e.g., assess surface roughness or detect microscopic contamination in a blood cell samples. From the beginning, the solution was required to facilitate the easy switching of camera technologies to optimize measurement time and resolution. Moreover, due to its extensive available post-processing capabilities, Python was selected as a software language. In the end, GenPyCam was employed both during the proof-of-concept phase to reduce development time significantly and on the final product to enable rapid switching between different cameras.

In-line inspection of textile using AI and an industrial IPC

#In-situ testing     #Reliability

Vision technology is increasingly being used for real-time quality control in production. GenPyCam facilitates a quick vision concept deployment and validation on industrial hardware. Recently, GenPyCam was used to develop and eventually deploy an in-line solution to detect weaving defects in textiles. Initially, it was unclear which technology (wavelength) could robustly identify faults. Therefore, GenPyCam was used to generate relevant datasets, allowing the training of AI models and, eventually, the evaluation of these technologies. Moreover, thanks to its quick programming and overall flexibility, GenPyCam proved crucial both in reducing the risk of potential downtime due to hardware failure and in the final deployment on an industrial IPC.

Easy sensor fusion: combined thermal and 3D vision

#Camera fusion     #Performance

Combining various camera viewpoints intelligently often yields information different from each individual camera viewpoint. Multiple RGB camera viewpoints allow, for example, the creation of a three-dimensional perception of the environment, which, combined with thermographic images, enables accurate analysis of the temperature of a particular object. GenPyCam has proven crucial in combining multiple cameras and camera technologies. The toolbox simplifies the hardware-software interface and allows the efficient and performant use of cameras and a structured project organization. The latter three are fundamental when many cameras are combined and deployed on a computer with limited computational capacity. Moreover, GenPyCam proved helpful during the proof-of-concept phase, allowing rapid testing of in-situ mixed-technology vision concepts to find the optimal problem-solution fit, resulting in highly robust solutions.