In virtual reality and augmented reality systems, predictive tracking technology can predict the next direction and/or position of an object or a part of the body. To give you an example, when your head is turning in a certain direction, you can synchronize to predict where your hand might be placed.
Why is predictable tracking technology so useful?
A common application of predictable tracking technology is to reduce the "motion photon" delay time, which is the time when the user completes an action in a virtual reality or augmented reality environment until the action is fully reflected on the display. . Since the motion itself will produce some delay, when the motion is completed and the information is displayed on the screen, there will be a delay (see below for the reason for the delay), so when you can predict the next direction and position, Then prioritize the relevant data to the screen to greatly reduce the perceived delay.
Although there are many concerns about predictable tracking technology in virtual reality applications, in fact, this technology is also very important in the field of augmented reality, especially when the user moves instantaneously in the real world, and the corresponding augmented reality overlay Reflected on the display. For example, if you bring an augmented reality helmet and see the enhanced display overlay on top of the solid object, more importantly, even if you rotate your own head library, the "coverage" in the augmented reality environment needs Still "locking" on the original object, rather than turning as your head turns, it actually shows a part of the real world. The object may be recognized by the camera, but the camera may take some time to capture the frame so that the processor can determine the position of the object in the frame, and then the graphics processing chip needs to re-render the new position on the overlay. By using predictable tracking technology, motion processing can be reduced in overlays compared to the real world.
How can predictable tracking technology work?
If you see a car driving at a constant speed, then you want to predict where the car might reach after the next second seems to be a very simple matter, and without any surprise, the accuracy of the prediction is also It will be very high - because you not only know the exact location of the car, but also know its current (or estimated) speed, so with these conditions, you can basically infer the future of the car. Where is the location.
Of course, if you want to predict where the car will be able to drive in a second, then this prediction may not be 100% accurate every time, because during this time, the car may change direction, or it may accelerate. So for car driving forecasts, the farther you try to predict, the more inaccurate the forecast will be. Simply put, if you let the car predict where the car is in a second, the accuracy is definitely higher than the position where you predicted the car after one minute.
In addition, the more you know about the car itself, the greater the chances of predicting accuracy. For example, you not only measure the speed of the car, but also measure its acceleration, and then make more accurate predictions.
Therefore, if you can get more behavior information of the object being tracked, you can also improve the prediction accuracy. By way of example, when you perform head tracking, if you can understand the speed of the head rotation and the angle of possible rotation, you can further optimize the tracking model to make a more accurate judgment. Similarly, if you are doing eye tracking, you can also use the acquired eye tracking information to predict head movement (this will be discussed in detail later in this article).
Cause of delay
VR helmet delay test device
The desire for predictable tracking technology comes mainly from the delay in virtual reality and augmented reality environments. Usually this delay refers to the action (or the effect of the action) after the user completes an action in the real world. How long does it take to reflect on the display? In fact, there are several reasons for delays, such as:
Sensory delay: Sensors (such as gyroscopes) may be bandwidth limited and will not report direction and position change information immediately. Similarly, camera-based sensors may also experience delays because the camera sensor receives light pixels from the tracking object and it takes a while until the corresponding frame is ready to be sent to the host processor, thus causing a delay.
Processing delay: Sensor data is usually integrated using some sensor fusion algorithms, and if the correlation algorithm is executed, there will also be a delay between receiving the data and outputting the answer to the algorithm.
To ensure smooth data, delays: sensor data is sometimes messy. In order to avoid errors, some software or hardware-based low-pass algorithms may be executed. Although the data fluency can be improved to a certain extent, it is because the algorithm quality is better. Low causes delay.
Transmission delay: For example, if a USB interconnect device is used for direction sensor data collection, it will take some time between the host processor to collect data and data to transfer data over USB, causing delays.
Rendering delay: When you are rendering a complex scene, it may take a while for the processor to determine (or decide) where each pixel in each frame is placed and when the frame is sent To the display.
Frame rate delay: For example, if you are using a 100Hz display, it means that the time from one frame to the next is about 10 milliseconds. When drawing a particular pixel, the current information may not be matched in real time because you need to wait for the pixel to be drawn onto the display to appear the next pixel.
Some of the reasons for the delays mentioned above may only lead to very small delays, but if several problems are added together, it may bring a bad user experience. In fact, predictable tracking techniques (in combination with other techniques such as time warping algorithms) are very helpful in reducing significant latency.
How far is the distance forecasting the future?
In fact, there seems to be no standard answer to the above question, because ultimately it depends on the user experience and feelings, and everyone's sense of body is also different. In the beginning, you may first estimate the end-to-end delay of the system and then optimize the time according to your preferences.
In addition, when forecasting for a given time period in the future, you may need to split this time into several forecast time periods. Why do you want to do this? We use the following examples to explain:
Different objects have different end-to-end delays. For example, camera-based hand tracking and camera-based head tracking show different delays, but in some virtual reality or augmented reality applications, you need to move your hands and head. Synchronized, so you need to use different predictive tracking times to make the final result more coordinated.
When configuring a single screen for a binocular image (such as a smartphone screen), there is also a half frame delay for a monocular image display (for example, half of a sixtieth of a second, or about 8 milliseconds). Only the eye will "see first" the image, and then the image will be reflected on the other eye. So in this case, it is best to use predictable tracking technology to judge 8 milliseconds in advance to reduce the half-screen delay experience.
Common prediction algorithm
Here are a few examples of predictable tracking algorithms:
Navigation speculation algorithm: This is actually a very simple algorithm. If the position and velocity (or angular position and angular velocity) are known at a given time, and the speed remains the same, the velocity value is not There will be errors, then you can calculate the final "foothold" of the object, and then carry out position prediction. For example, if the last known position is 100 units and the last known speed is 10 units per second, then the position of the object is 100+10*0.01=100 during the next 10 milliseconds. 1. Of course, this is a very simple calculation because it has a premise that the position and speed are correct (and subjectively not affected by other measurement problems), and the speed is always constant. But in fact, all of these assumptions are almost impossible to achieve in reality.
Kalman predictor: It is based on the well-known Kalman filter used to reduce system sensor noise in mathematical models of existing system operations. If you need more information about the Kalman filter, please see here.
Alpha-beta-gamma: The ABG predictor is highly correlated with the Kalman predictor mentioned above. Although the mathematical application is simpler, its versatility is more general. The ABG predictor attempts to continuously estimate velocity and acceleration and then applies this data to the prediction. Since the valuation takes into account the actual data, they reduce some of the valuation errors. Configuring alpha, beta, and gamma parameters also enhances responsiveness.
Predictable tracking technology is a very useful virtual reality/augmented reality technology, and it is also used to reduce performance delay. Although the implementation of this technology is simple, it requires a high level of professionalism, so you need to do some thinking and analysis, because in the current virtual reality and augmented reality systems, reducing the delay is crucial.
BLPS laser safety protective device is designed for personal safety used on hydraulic bender.
The dynamic test technology it used has passed the Type 4 functional safety assessment by TUV, and get the national invention patent. The product reaches the advanced technological level of similar products.
BLPS laser safety device provides protection zone near the die tip of the bender to protect fingers and arms of the operator in close to the upper mold die tip. It is the most effective solution so far to preserves the safety and productivity of the bender.
Press Brake Protection,Laser Guarding Device,Press Brake Guarding Systems,Press Brake Guarding
Jining KeLi Photoelectronic Industrial Co.,Ltd , https://www.sdkelien.com