- Standalone solution for storing acquired images
- Easy retrofitting of existing image processing systems for storing acquired image data
- Wide range of applications: automated, economical classification of image data for deep learning tasks
- Fast replacement of SSD data storage devices if required
- Compact, robust design with passive cooling for use in industrial applications
The Image Logger is a stand-alone solution for storing captured images, designed to operate in parallel with a vision system. This product is ideally suited for retrofitting into existing systems and greatly simplifies, for example, the training of Deep Learning systems or the documentation of image evaluations.
Many image processing systems currently in use are not designed to permanently store captured images. In numerous applications, however, image data storage is a useful addition to existing image processing systems and expands their capabilities enormously. For example, system setup and troubleshooting are greatly simplified. In addition, the Image Logger enables the documentation of quality inspections to meet legal requirements or also the more efficient training of a neural network.
The Image Logger offers an effective way to store image data in addition to existing systems, even retroactively and without modifying the existing vision system. The setup is extremely simple: The Image Logger is connected via prepared Ethernet interfaces between the GigE cameras used and the existing image processing computer. Apart from an additional Ethernet cable, no further hardware is required, as the Image Logger is connected directly between the image processing system and the GigE camera. There is no need to develop software for the Image Logger; it operates completely autonomously.
In operation with a new setup, the image data is acquired as before via the existing camera or, in the case of multi-camera systems, via several cameras, but now it is first transferred to the Image Logger, where it is stored in RAW, BMP or JPEG data format with a time stamp. From there, the Image Logger forwards the image data to the existing image processing computer. This takes over the data processing in unchanged form and returns the results of its evaluation to the system as before, for example to trigger the sorting out of products that do not meet the quality criteria. The N.O.K. signal can be routed through the Image Logger, which in this way also allows the results of each N.O.K. image evaluation to be saved.
This feature opens up interesting technical possibilities. For example, companies that need to document the results of their quality assurance for each individual product, for liability reasons, can solve this task easily and reliably with the help of the Image Logger. The integrated SSD image memory can be replaced at regular intervals without much effort. Via Ethernet or by connecting the SSD to a Windows or Linux PC, the stored images can be retrieved for subsequent documentation without any additional software.
For the increasingly important topic of Deep Learning, the Image Logger provides another outstanding capability: Since the stored images including their classification as e.g. O.K. or N.O.K. objects are available in the system, this already labeled data can be used very easily as a basis for training, for example, a neural network. In contrast to the otherwise usual and time-consuming approach, in which each image must be viewed and classified by a suitable expert, the Image Logger thus offers an extremely economical way of making existing image data usable for Deep Learning applications.
The Image Logger can be used in a wide range of system architectures: it is suitable both for simple image processing systems with just one camera and one PC and for more extensive machine vision systems in which several cameras are connected to one or more evaluation computers. The experts at IMAGO Technologies will be happy to support you in the optimal expansion of your system with one or more image loggers.
Applications whose image processing systems have to be retrofitted for storing image data after a certain period of use can be found in almost all branches of industry. The following examples show some possibilities of the Image Logger.
Food & Beverage
Today, automated production of food and beverages is increasingly taking place at enormous process speeds. The Image Logger expands the possibilities of image processing systems that reliably inspect quality features at high speed and capture codes or best-before dates and store them with reliable temporal assignment. In particular, image storage for the feasibility of deep learning solutions will play an increasingly important role in the future.
In the pharmaceutical industry, there is a legal requirement that details of the production process, such as the exact date of manufacture or the exact batch of each drug, must be seamlessly traceable to the end user. Only in this way it is possible to reliably carry out recalls when faulty medications occur, thus minimizing the risk to patients’ health. By storing images during drug production and allowing quality characteristics to be reliably documented and tracked retrospectively, the Image Logger offers drug manufacturers an optimal way to meet the requirements for documenting their products.
Automated quality inspection of objects in industrial production represents one of the main application areas of machine vision due to the extremely diverse processes and products. In this context, many of the machine vision systems in use have already been in operation for years and do not have the necessary flexibility to meet additional requirements when switching to new products in terms of defect documentation. Retrofitting such systems with the Image Logger is usually more economical than replacing the entire machine vision system.
In addition, the Image Logger simplifies troubleshooting in production sequences and thus contributes to process optimization. When converting image processing systems from traditional algorithms to methods from the field of Deep Learning, the Image Logger also offers extremely economical options for using already existing and classified image data for training neural networks without the need for cost-intensive specialists. Due to its compact design and robust construction with passive cooling, the Image Logger is ideally suited for use in industrial applications.
It is very important for manufacturers of electronic components to be able to uniquely identify their products at any time in order to avoid? economic damage due to product piracy or unjustified support claims. With the help of the reliable documentation of unambiguous component markings by the Image Logger, such misuse can be ruled out. The doubtless proof of defect-free products at the time of delivery also reduces the economic risks for companies in electronics production.
Packaging & Logistics
Were the ordered goods addressed to the correct recipient? Was the packaging flawless at the time of shipment? Were all items picked correctly according to the order? Questions like these can only be answered with certainty in retrospect if logistics processes are reliably documented and clear proof can be provided in case of doubt. The Image Logger offers companies in the packaging & logistics industry a wide range of options for such tasks.
In the printing industry, high production speeds and frequently changing products and motifs are commonplace. Especially in flexo printing, manufacturers often need to be able to prove that their goods have been produced faultlessly and as required by the customer long after production. This can be reliably documented with the help of the Image Logger.
The Image Logger offers machine builders as well as manufacturers of inspection or production machines of all kinds a simple option to retrofit existing image processing systems in such a way that they allow reliable storage of image data. Based on this data, reliable documentation of product quality is possible, among other things. Users looking for economical ways to deploy innovative Deep Learning applications will find Image Logger an effective tool for automated classification of image data for subsequent Deep Learning tasks. The intuitive web-based user interface of the Image Logger greatly facilitates the use of this innovative product for users in all industry sectors.