Smart cameras for cloud computing

The journey to cloud based smart cameras

Smart cameras have now established themselves as stand-alone systems in various applications for the evaluation of production parts. Such cameras offer a particular advantage due to their compact, powerful design in areas where no additional computers or their peripherals can be installed. The disadvantage is that the cameras are usually designed for only one area of application and must be constantly adapted as production processes change. Therefore, the lifetime of smart cameras in these areas often does not meet the expectations of the users. The need for adaptable, intelligent systems that can be maintained despite possible changes and expansions in production lines and that can grow with technical innovations becomes apparent.

The journey to Cloud-Computing

For this reason, the solution trend is emerging that smart camera systems are increasingly being controlled via AI applications and cloud computing. In the process, the advantages of both systems are merging. On the one hand, the smart cameras themselves offer an extensive possibility for pre-processing images, whereby the scalable performance strength of external computer units in particular can provide an optimized solution to the increasing demand for difficult verification methods. At the same time, this results in the possibility of better synchronizing the systems in their entirety with the plant. Many users are currently looking for a bridging technology, which means that they would like to solve their tasks classically with a smart camera for the time being, but still be prepared for a later switch to cloud computing. Ideally, no new hardware should be required for this and the camera should only be adapted to the new environment by means of software updates. In this way, little money can be spent on peripherals and infrastructure and yet a long-lasting concept can be integrated.

evoVIU as cloud based Smart Camera

The evoVIU can meet both requirements, as it can be used both as a stand-alone smart camera and as a cloud-based smart camera. And it does so without any necessary adjustments to existing hardware conditions. This offers users obvious advantages. The evoVIU can be integrated quickly, requiring only a power supply and an Ethernet line. Afterwards, the camera can be set up directly on site and without installing additional software for individual circumstances via all end devices.

When using the smart camera without cloud-based solutions, image processing tasks can be performed directly on the camera and sent to defined endpoints. If the evoVIU smart camera is to be controlled via the cloud, commands for controlling and regulating image acquisition can be provided via an encrypted http REST interface.

Workflow of the evoVIU for cloud-based use

After the evoVIU has been successfully installed, the optimal setting of the desired image follows. Once these two points have been completed, the UUID uniquely created during this process can be saved and transferred to the AI programmer. He writes the code for evaluating the captured part, saves the image and thus optimizes the training set. When production and acquisition start, the image source is created and activated on the evoVIU. After a trigger input, the image is captured and processed. The processed image is evaluated, classified as ok or not ok and stored for further reference and optimization of the training set.