Abstract: A method for processing video data for increased sharpness. The data undergoes a process by which it is filtered in two dimensions. Using separable, one-dimensional filters, the process cores the data to prevent noise enhancement and applies gain to the resulting data. Finally, both dimensional filtered components are then combined to form the final sharpened image. Because of the separable nature of the process and filters, either one of the dimensional filters could be eliminated, while retaining the robustness of the method.

Abstract: An improved method of line screen printing. Instead of screening the image with only one set of lines, the image is also screened with a second set of lines, which are substantially orthogonal to those of the first set. (FIG. 3). The screening is accomplished with two types of cells, horizonal cells and vertical cells. The pixels are classified according to their distance from the nearest line in the horizontal cells. This process is repeated with respect to the vertical cells. (FIG. 5). Different tone curves, which map input pixel values to output pixel values, are associated with different pixel classifications. In general, pixels closer to lines have tone curves that map to higher output values. (FIGS. 6 and 7). Because each pixel is in both a horizontal cell and a vertical cell, each pixel has two values from the tone curves. For each pixel, its two tone curve values are combined to obtain a greyscale value for that pixel.

Abstract: Methods of reducing artifacts in SLM-based display systems (10, 20), whose images are based on data displayed by bit-weight for pulse-width modulated intensity levels. A first method is suitable for systems (20) that use multiple SLMs (14) to concurrently display images of different colors, which are combined at the image plane. The data for each color are staggered in time (FIG. 4). A second method is suitable for either multiple SLM systems (20) or for systems (10) that use a single SLM (14) and a color wheel (17) to display differently colored images sequentially. The data for each color is arranged in a different data sequence (FIG. 5). In either method, the intensity transitions do not occur at the same time.

Abstract: A color wheel (15) for use in a display system (10). The color wheel (15) is comprised of a rigid hub (23) and an outer perimeter of color filter segments (21), which are made from a thin plastic material.

Abstract: An architecture for a compact, 1.times.N optical switch. The switch package receives light from an input optical fiber (12b), which is directed over a well or gap at the bottom of which lies a micromechanical structure (10). If the structure is in an unaddressed state, the light travels into an in line output optical fiber (12a). If the structure (10) is in an addressed state, it intercepts the light and reflects it out of the plane of the input optical fiber to an offset mirror (24). The offset mirror (24) then reflects the light to one output fiber (16a, 16b). The offset mirror may have steps such that more than one optical fiber could become the output fiber, depending upon the structure's position.

Abstract: A membrane device is provided in which a plurality of ridges (14) and recesses (16) are formed proximate to a substrate (12). Electrodes (18) are formed within the recesses (16). A spacer (20) supports a membrane (22). Application of a potential difference between the membrane (22) and the electrodes (18) allows for deflection of the membrane (22) toward the electrodes (18).

Abstract: Higher quality printing is difficult in implementation in spatial light modulator printers. The two major problems are accomplishing gray scale within the line time constraints, and eliminating staircasing artifacts within the images printed (81). It can be improved by using an alternate way of resetting cells on the spatial light modulator when data is being loaded onto the cells, timing delay (86), horizontal offset (84), and differently sized pixels (80, 82).

Abstract: An improved process for manufacturing semiconductor devices (10). The device, which could be a semiconductor wafer, an individual chip, or a device that has integrated within it semiconductor devices, such as a compact disk drive, is placed or held upside down with its working surface (12) open. The device (10) is struck, causing particulates (18) attached to the working surface (12) to fall free of the device. Alternately, the device could be sharply decelerated to apply the shock. Additionally, the device could be held substantially vertical and rotated to use centrifugal force to separate the particulates away from the device.

Abstract: A method for causing a micromechanical spatial light modulator to display data for a period less than its settling time. The modulator elements receive a first pulse (40) that causes them to release from their previous state, a bias voltage is removed and reapplied, allowing the elements to move to the unaddressed state, and then the elements receive a second pulse (46). After receiving a second pulse, the elements assume an unaddressed state. In one embodiment, new address data is loaded during this unaddressed state, after which a bias is reapplied causing them to achieve the state corresponding to the new state. In another embodiment, the previous addresses are cleared during the unaddressed state, forcing the elements into an OFF state. In either embodiment, a reset pulse may be applied after either the load or clear step.

Abstract: A method of pulse width modulation using a spatial light modulator (40) with a finite transition time. The method uses m bits per sample to digitize the incoming data, but apportions the LSB times for pulse width modulation based upon m-1 bits. The current video frame displays all of the bits for each sample, except for the LSBs for each sample. The next video frame displays all of the bits for each sample, adding one more LSB for dividing up the frame time. The first frame could use either the additional LSB time and display no data, or it could use only that number of LSB times it needs. In the latter, the system will have to adjust to different partitions of the frame time for alternating frames. The system includes a spatial light modulator (40), a memory (42), a formatter (48), a sequence controller (44) and a toggle circuit (46), to perform this method.

Abstract: A line generator (31) for receiving fields of pixel data sampled from a video input signal and for generating additional lines of pixel data so that the display frames will have more lines than the fields. The line generator (31) has a motion detector (31a) that determines, on a pixel by pixel basis, whether some part of the current field is in motion. A motion signal from the motion detector (31a) is used to select between outputs of two or more pixel generators (31b, 31c). One of the pixel generators (31b) provides pixel values that are better suited for display when the image is not in motion. The other pixel generator (31c) provides pixel values that are better suited for display when the image is in motion.

Abstract: A non-linear torsion hinge (12, 22) for a micro-mechanical device (10, 20) having a hinged movable element (11, 21). Each hinge (22) is comprised of two hinge strips (22a) spaced apart in the same plane, such that the axis of rotation of at least one of the hinge strips (22a) is different from the axis of rotation of the movable element (21). As a result, the hinge strip (22a) must elongate as it twists, thereby providing a greater restoring torque.

Abstract: Methods of reducing artifacts in SLM-based display systems (10, 20), whose images are based on data displayed by bit-weight for pulse-width modulated intensity levels. A first method is suitable for systems (20) that use multiple SLMs (14) to concurrently display images of different colors, which are combined at the image plane. The data for each color are staggered in time (FIG. 4). A second method is suitable for either multiple SLM systems (20) or for systems (10) that use a single SLM (14) and a color wheel (17) to display differently colored images sequentially. The data for each color is arranged in a different data sequence (FIG. 5). In either method, the intensity transitions do not occur at the same time.

Abstract: Higher quality printing is difficult in implementation in spatial light modulator printers. The two major problems are accomplishing gray scale within the line time constraints, and eliminating staircasing artifacts within the images printed (81). It can be improved by using an alternate way of resetting cells on the spatial light modulator when data is being loaded onto the cells, timing delay (86), horizontal offset (84), and differently sized pixels (80, 82).

Abstract: A method and structure for a display system having multiple spatial light modulators (SLMs) (16), each of which contributes an image of one color that is perceived by the viewer as a combined image. The SLMs (16) have more rows and columns of pixel elements (42) than rows or columns of pixel data to be displayed. A window of "active" pixel elements (42) can be shifted up and down or right and left by selecting which pixel elements (42) are to receive data. The addressing circuit (31, 31a, 35, 35a) of each SLM 16 can be controlled so as to accomplish this shifting.

Abstract: A method for printing grayscale images that compensates for problems in the printing process. The presence of a defective element on a spatial light modulator (12) is compensated for using a background value. The background value is determined by the number of defective elements stuck in the ON position, and that value is added to all of the exposure values of the elements, creating exposure data. In one embodiment, a defect intensity value is subtracted from those pixel exposure values affected by defective elements. The exposure data is used to generate microimages on a photoreceptive drum, the data is decremented and the steps are repeated until an entire image has been produced.

Abstract: A method for printing grayscale images that compensates for problems in the printing process. The presence of a defective element on a spatial light modulator (12) is compensated for with use of a white balancing area. If a column of the modulator (12) has an element stuck ON, a corresponding element in the same column in the white balancing area is turned OFF while all other columns have elements that are ON. If the illumination is non-uniform, the exposure values of each element are multiplied by an efficiency factor which compensates for the non-uniformity.

Abstract: An optical guide (10, 40, 50, 60) for horizontally aligning two vertically stacked images generated by one or two SLMs. The optical guide has a channel separator (10a) that directs both images along two different paths. A pair of aligning reflectors (10b and 10c) on each path vertically shift the images with respect to each other so that at least part of the images on the first path are aligned side-by-side with at least part of the images on the second path. The channel separator (10a) then re-directs the images to the image plane 15. Along both paths, at least two of the reflecting surfaces of channel separator (10a) or aligning reflectors (10b and 10c) are optically powered so as to change the width or height of the images.

Abstract: A method of providing control signals for resetting mirror elements (10,20) of a digital micro-mirror device (DMD) having reset groups (FIG. 4), or for resetting moveable elements of other micro-mechanical devices that operate with similar principles. A bias voltage is applied to the mirrors and their landing sites, and an address voltage is applied under the mirrors. (FIG. 3). The address voltage is held at an intermediate level except during a reset period. During this reset period, the address voltage is increased. Also, during reset, the bias applied to mirrors to be reset is pulsed and offset, and the bias applied to mirrors not to be reset is increased. (FIGS. 5 and 6).

Abstract: An improved reticle (20) and method of using it to expose layers of wafers for large integrated circuits (10). The integrated circuit (10) is designed so that nonrepeating patterns are laid out in perimeter areas, distinct from the center area containing contiguous repeating patterns. The reticle (20) is patterned with multiple masks (21-23), with different masks representing the repeating and nonrepeating patterns. The mask (22) representing the repeating pattern may then be stepped and illuminated separately from any mask (21, 23) representing a nonrepeating pattern.