High-speed recording solutions
Often the frame rates required are 10 to 100 times faster than taking one picture per cycle as in on-line systems. This type of application records high-speed image sequences both before and after an event enabling detailed off-line analysis. Fast frame rates of up to a few thousand frames per second can be achieved, which is fast enough to capture a wide variety of potential manufacturing process problems.
High-speed image sequence recording is a diverse technology used in many application areas including vehicle crash testing, high-speed process and machinery diagnostics and for recording and evaluating complex procedures in medical applications. It is also used for statistical evaluation in sports, and for high-speed action scenes in film making.
There are two types of high-speed recording solutions, on-line and offline. On-line systems rely on high-speed cameras sending data directly to a PC's system memory or disk capacity, while off-line solutions contain the image recording memory within the camera itself. Both solutions have advantages and disadvantages and the decision is normally driven by the required frame rate.
On-line systems rely on the interface speed between the camera and the computer and the speed of the computer. Using the fastest current interfacing technology (status 01/2015) this is limited to roughly 4 GB/s. If this speed is acceptable and you require long sequences this solution is a good choice.
Off-line systems increase this throughput as images are held in the camera without any need for transmission while recording, however, these systems are normally limited to 16 GB or 32 GB of memory, reducing the maximum possible recording length.
A new approach for a different storage concept using a solid-state storage device (SSD) in the camera allows image data rates to increase massively. This means that high frame rates combined with high resolution for high-performance slow-motion and high-speed cameras are now possible. This new data management provides camera-internal data transfer at 1300 MB/s. That is six times the data speed compared to conventional storage concepts.
Triggering and synchronisation
In on-line inspection systems the goal is to capture a single image at the trigger point when the object passes the camera. Generally there is no concept of a sequence in this type of application. For sequence recorders the trigger point is the point that starts recording the whole sequence at full speed. Of course it is also possible to synchronise each frame within the sequence to the process.
Types of sequence trigger
As there are many applications for sequence systems, many different requirements for triggering scenarios exist. The simplest method is to start recording when a trigger event occurs. The drawback here is that the recording of the information cannot occur before the trigger event, which is a common requirement in diagnostic use. To achieve this, recorders use the concept of a ring buffer. In this mode, the system continuously records into memory. Once full, the system starts overwriting the first records in memory.
On a trigger event, the system records until the ring buffer is filled up to a certain position and stops, thus providing, both pre- and post-event recording information.
Sequence recorders sometimes offer other methods of triggering than the traditional trigger input. One such method is using image content. If the image data in a certain area changes significantly, either in intensity or because there has been movement in the image, the camera triggers itself to send an image or sequence, removing the need to generate a trigger using hardware. In this example the sensor automatically triggers an image whenever a car passes the trigger activation field. The IR illumination allows an image to be captured under any conditions.
This diagram shows a set-up in which a vehicle test collision is being monitored to collect data about the event. An area of the scene is being monitored by the camera and used to trigger the recording sequence if anything within that area changes according to predetermined values. This provides a trigger exactly at the moment required, without the need for complicated external sensors and triggering.