Computer graphics processing and selective visual display system – Computer graphics processing – Adjusting level of detail
Reexamination Certificate
2000-11-07
2003-04-15
Jankus, Almis R. (Department: 2671)
Computer graphics processing and selective visual display system
Computer graphics processing
Adjusting level of detail
Reexamination Certificate
active
06549204
ABSTRACT:
TECHNICAL FIELD
The present invention relates to graphic image generation. In particular, the present invention relates to the generation of models for use in training simulators and the like. More specifically, the present invention relates to the generation of graphic images in which certain objects, whose area is a small percentage of the total image and yet of critical importance to the utility of the interactive graphic task being performed, are generated prior to use in the simulator and are provided with enhanced features to accurately represent the object.
BACKGROUND ART
High quality graphic image generation is used in various fields such as flight simulators, medical training and surgery, computer games, and engineering workstations, to name a few. It is imperative that these systems provide realistic images for the benefit of the user. These images should have as a minimum sufficient quality to correspond to the visual scene experienced by the user in viewing the objects directly with either optically aided or non-aided vision. The overall objective is to facilitate the teaching or game playing environment for the benefit of the user. The system goal therefore is to provide an immersive environment which is perceived by the user to be very like the visual appearance of the task as it would be performed in the real world.
The versatility of computers and emerging graphics display technology has led to the development of computer based training in which a range of tasks are presented through the visual medium. As may be expected, the closer the graphics are to those encountered in real situations, the higher the confidence level of the student and instructor in the value of the training conducted.
The generation of a graphic image by a computer relies on the existence of a model representation of the object which is being imaged. This model precisely defines the geometric and visualization properties of the object for graphic presentation. Depending on the view which is desired, the visualization process transforms the inherently 3-Dimensional representation into a two dimensional object. Appearance of the representation mimics that of the actual object as though it were being viewed through a virtual “window” whose position corresponds to that of the display medium.
The display medium itself presents limitations on the display of objects. The standard method of rendering an object is to transform it into a series of discrete picture elements (“pixels”) which constitute the image of the object. Since these pixels have a very well defined and static size on the display, they impose a limit on the accuracy and detail of the image. This represents a distortion of the objects' model and detracts from the realism of the display. The standard industry answer to this problem has been a continuing effort to decrease the size of the pixels so as to improve realism.
A second problem encountered is that the complexity of the visual scene may lead to poor performance by the graphics system. This either reduces the allowable richness of the visual environment—to maintain realistic performance—or causes the expense associated with graphics hardware to rise, to allow performance expectations to be met. The increasing speed of graphics chips is steadily driving the price of graphics systems down and enlarging the scope of applications which can be handled by training systems.
The limitations of the display constitute the problem to be addressed. The following kinds of problems are present in known displays:
1. Aliasing: The image appears to be composed of “blocks” of a given size. Lines appear to be staircases, circles have sawtooth boundaries, etc.
2. Shape Distortion: The finite size of the pixels causes small objects to be represented by single pixels in which the aspect of the object cannot be determined.
3. Accuracy: The intensity of the object is not correctly computed and the object flickers due to random subsampling of the small object as successive frames are computed.
4. Special Effects: A lack of cues that the user may specifically rely on to assist in identifying or tracking the motion of an object. For example, sun glint from a windshield canopy.
A basic assumption is that only certain, very specific objects need to be rendered with high accuracy and detail. If an attempt were made to render the entire scene with additional accuracy, the graphic processor would be swamped with the requirement and the potential advantages of an improved rendering would not be available.
To this end, it is desired to provide systems which do not create false impressions with unrealistic or inaccurate object representations. For example, flight simulators are employed to train fighter pilots on how to quickly detect objects such as enemy planes and missiles. The pilot does this by scanning the horizon in a predetermined pattern along with other visual and auditory warnings. As such, if the flight simulator renders an object with a fluttering appearance or an unrealistic large size, a false impression of the target is generated. As such, the training exercise is detrimental in that the actual appearance of an enemy plane or target is unrealistic. Hence the visual expectations of the pilot in air combat become unrealistic and life threatening. Similar limitations exist for medical training and surgery preparation/execution and other similar applications. Hence accepted practice is to overcome this very serious limitation by resorting to alternate—and generally more expensive—means of accomplishing the objective. As an example, after training in a simulator, the pilot must spend a large number of hours in the aircraft to become familiarized with the appearance of aircraft and missiles in the real world.
Attempts at improving graphic image processors used with simulators and other interactive graphics devices continue due to the desire to improve the quality of displays. These are limited by the pace of advancements in computer and display technology.
In general, it is known to apply improved rendering techniques to an entire image display to enhance the overall appearance of the images presented. However, this approach rapidly consumes processing power available and accordingly, limits other operational aspects of the image processor such as real-time presentation of the total visual environment. Moreover, current technology graphic processors using embedded graphic algorithms are unable to selectively improve the visual appearance of those items whose detail is particularly important and critical to the overall success of the training simulation. This is exemplified by the aforementioned planes and missiles that require high acuity presentation in order to assure that the pilot is being trained in an environment as similar as possible to the visual environment likely to be encountered in actual air combat.
The present processing equipment does not prioritize these objects and accordingly, processes the important items as it would any background information. This limits the usefulness of the training or display environment.
One alternative to the aforementioned approach is to employ high acuity projectors in conjunction with a graphic image processor. This technique generates a simulated background scene and superimposes the critical images onto the scene with a higher resolution. This requires additional processing equipment and is quite expensive. Moreover, the high resolution projectors of today are unable to represent the critical objects with the acuity and real world appearance necessary for effective training or practice.
An extension of this approach is to provide a hardware-based solution utilizing high resolution Area of Interest displays. In conjunction with this, a mechanism is provided for tracking pilot head position and those areas where the pilot is perceived to be looking are processed with high resolution. Unfortunately, this method employs unrealistic background scenes which appear artificial and do not present an accurate representation for a training simulator. Hence
Jankus Almis R.
Renner Kenner Grieve Bobak Taylor & Weber
LandOfFree
Intelligent model library for a graphic image processing system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Intelligent model library for a graphic image processing system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Intelligent model library for a graphic image processing system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3086723