REAL TIME RECORDING FOR VISION APPLICATIONS

Machine Vision Application Areas

Machine vision and/or Computer Vision is mostly involved in processing and analyzing video directly from the camera (with any resolution or frame rate) to attain some results based on the project. Most of the vision applications are to INSPECT quality in different terms according to the objects and projects. These results are mathematical, or signal output to drive next event. For example, in a manufacturing factory, a product or part of the product made is inspected for its features in terms of shape, size, placement, quality, positioning, barcodes, OCRs, patterns etc., and the results in terms of measurements, images obtained and if needed the following mechanism like Conveyor, Pick and place ROBOT or any other PLC driven machinery are provided with the I/O signals, values to PLC etc.

Machine Vision

Mostly Frames generated by Machine Vision Cameras are not recorded as Video for any use unless it is for Prototyping. But these machine vision cameras have got its usage expanded into the field of Science Research, Microscopy, Machine Break Down Analysis, offline analysis of fast events and so on. Some of these applications even demand multiple synchronized camera usage for coordinated analysis. These usage and applications drive the imaging field to work on Videos and also Uncompressed Video.

 

Now, we get into a term uncompressed. As the camera markets confuse customers these days with abundance of IP Camera manufacturers, Machine Vision Camera manufacturers, Microscopy camera manufacturers, Scientific Camera manufacturers for PIV and other applications. Uncompressed in simpler term means – just get the RAW and undisturbed video to the system where analysis to be done. IP Cameras always do compression within. Why? It is simple because of the interface used. Most of the USB 2.0 or Ethernet based Cameras will have to do compression to accommodate the high resolution and better frame rates to be transported to the PC system or recording device. Machine Vision Cameras which are more used for processing images or frames are always with a feature of providing Uncompressed Video as output. They come in different interfaces like GigE (Vision), USB3 (Vision) and more complex 5G, 10G, Camera link, Coax press, HD SDI etc.

 

In general, all these Machine Vision Cameras get shipped with a Camera configurator software and SDK software to work with those cameras in terms of setting parameters, capturing frames, trigger, I/O controls, Display etc. These packages will provide the software features where one can capture images/frames as video to the memory of the Computer to which the camera is connected. Few manufacturers provide a tool to record video on to HDD to a certain extent but not optimized and will land up in Frame loss etc. These tools are for simple Microscopy applications where few frames are needed to be seen later or processed later. These software will not have any thing optimized to do recording of Video to the HDD as per need. They work better for recording on Memory of the system and then transfer them to either frames or Video again in Uncompressed format.

High Speed Cameras

There are High Speed Cameras that have built in memory or SSD Hard disks which records video on the same and then stream it to the system where process has to happen. All these cameras are expensive and cannot be used if both live processing and recording has to be done as they always record and stream.

The need for using Machine Vision Cameras in uncompressed video recording on to a system or specialized recording system gave way to technology that emerged for the optimized use of memory, HDD, and OS to record video to the HDD on a PC or to a specialized recorder. Companies like NORPIX and IO INDUSTRIES from Canada specialize in these.

The recording software will basically store the video on memory or in HDD uncompressed but in a native format of their own as sequence. This sequence then can be transported to any format as AVI (the standard format) or use some CODECs and create MP4 or other formats or just break frames as Images in different formats like BMP, TIFF, JPG etc. The trick of the sequence recording in uncompressed format allows these frames available for storing either as video or Images separately. They effectively use memory maps and HDD synchronization techniques. The bottle neck would be to have right kind of HDD and ample memory and proper OS to achieve the needed recording.

In usage of Machine Vision Cameras, the uncompressed, Loss less Frame output are the key for applications. To achieve this, it is essential to have Professional software like Streampix from Norpix or DVRs from IOIndustries.

Bundled software from Camera companies may be good enough to record a video within the memory of the computer that too for very few seconds and based on memory availability. The video from memory can be transferred  offline to the HDD but when it comes to direct HDD recording then it is for sure they get into frame loss and also 2GB kind of file restrictions. These software do not conduct any study of the system where the camera is connected and just tries recording with many restrictions which may be referred in the manuals.

Hence, any application that needs video (loss less in terms of uncompressed and Frames), it is recommended to use a professional software built for the purpose. Moreover, these software or DVRs would have provisions to ensure the frames based on reference time with different inputs either through software like system clock or other devices connected to the PC or DVR.  Only these software and/or DVRs provide recording for minutes and even hours depending on the need and also system.

It becomes so important to understand that though we call as Video, Frames or images are so important for any analysis. These definitely need the professional software where companies have spent more than two decades to make such system for every type of OS released on PCs and also utilizing all hardware developments in the course of time.

Let us have a proper recording done on systems the output from Vision Cameras and have a peaceful system for proper and correct analysis.

Note: The above article is a collection from various web sites, manufacturers and distributors / integrator. If any company/individual finds any ownership of contents – they can let us know at info@onlsol.com to either change or remove the contents. Article is provided only for the information purposes and facts can be cross verified by the readers.

Recent Posts

Recent Comments

    Archives

    Categories

    Meta

    admin Written by:

    Be First to Comment

    Leave a Reply

    Your email address will not be published. Required fields are marked *