CN1981323A - Optical positioning device using different combinations of interlaced photosensitive elements - Google Patents

Abstract

One embodiment relates to an optical displacement sensor for sensing relative movement between a data input device and a surface (304) by determining displacement of optical properties in successive frames on the surface. The sensor includes at least a detector, a first circuit, and a second circuit. The detector includes a plurality of photosensitive elements organized into first and second arrays (e.g., 1502 and 1504). The first circuit is configured to combine signals from every M-th element of the first array to generate an M-group signal, and the second circuit is configured to combine signals from every M'-th element of the second array to generate an M'-group signal. M and M' are different numbers from each other. Other embodiments are also disclosed.

Description

Optical positioning device using different combinations of interleaved photosensitive elements

Related Application Cross Reference

This application claims the title "Optical position sensing device having a detector", filed May 21, 2004, by inventors David A. LeHoty, Douglas A. Webb, Charles B. Roxlo, Clinton B. Carlisle and Jahja I. Trisnadi benefit of U.S. Provisional Application No. 60/573,075 for array using different combinations of shared interlaced photosensitive elements". The disclosures of the aforementioned US provisional applications are incorporated herein by reference in their entirety.

technical field

The present invention relates generally to optical positioning devices (OPDs), and methods of sensing movement using such devices.

Background technique

Pointing devices, such as computer mice or trackballs, are used to enter data into and interface with personal computers and workstations. This device allows rapid repositioning of the cursor on the monitor and is useful in many text, database and graphics programs. The user controls the cursor by moving the mouse over a surface, causing the cursor to move in a direction and distance proportional to the movement of the mouse. Alternatively, movement of the hand on the stationary device can also be used for the same purpose.

Computer mice come in both optical and mechanical types. Mechanical mice typically use a spinning ball to detect movement, and a pair of shaft encoders make contact with the ball to produce digital signals that are used by the computer to move the cursor. One problem with mechanical mice is that they are prone to inaccuracy and malfunction after continued use due to, for example, dirt buildup. Furthermore, movement and combined wear of the mechanical elements, especially the shaft encoder, necessarily limits the useful life of the device.

One solution to the aforementioned problems of mechanical mice has been the development of optical mice. Optical mice have become popular because they are more robust and offer better pointing precision.

The main conventional technology for optical mice relies on light-emitting diodes (LEDs) illuminating the surface at or near tangential incidence, a two-dimensional CMOS (complementary metal oxide semiconductor) detector that captures a composite image, and correlating successive images Software to determine the direction, distance, and speed that the mouse has moved. This technique usually provides high accuracy but suffers from complex design and relatively high image processing requirements. Furthermore, the optical efficiency is low due to the tangential incidence of the illumination.

Another approach is to use a one-dimensional array of light sensors or detectors, such as photodiodes. Successive images of the surface are captured by imaging optics, converted to photodiodes, and compared to detect mouse movement. Photodiodes can be directly connected into groups with wires for easy motion detection. This reduces photodiode requirements and enables fast analog processing. An example of such a mouse is disclosed in US Patent No. 5,907,152 to Dandliker et al.

The mouse disclosed in the Dandliker et al. patent also differs from standard technology in that it uses a coherent light source, such as a laser. Light from a coherent source scatters off the rough surface to produce a random intensity distribution of light, called speckle. There are several advantages to using a speckle-based pattern, including efficient laser-based light production and high-contrast images even under normal incident illumination. This allows for more efficient systems and saves current consumption, which is beneficial in wireless applications to extend battery life.

While a significant improvement over conventional LED-based optical mice, these speckle-based devices are still not entirely satisfactory for a number of reasons. In particular, mice using laser spots do not exhibit the level of precision typically required of state-of-the-art mice, typically requiring path errors of less than 0.5% or thereabouts.

This disclosure discusses and provides solutions to various problems with prior art optical mice and other similar optical pointing devices.

Contents of the invention

One embodiment relates to an optical displacement sensor for sensing relative movement between a data input device and a surface by determining displacement of an optical property in successive frames of the surface. The sensor includes at least a detector, a first circuit and a second circuit. The detector includes a plurality of photosensitive elements organized in first and second arrays. The first circuit is configured to combine signals from every Mth element of the first array to generate M groups of signals, and the second circuit is configured to combine signals from every M'th element of the second array to generate M' groups group signal. M and M' are different numbers from each other.

Another embodiment relates to a method of sensing movement of a data input device over a surface using an optical displacement sensor having a detector comprising a plurality of light sensitive elements organized into first and second arrays. A plurality of photosensitive elements receive intensity patterns generated by light reflected from portions of the surface. Signals from every Mth element of the first array are combined to generate M group signals, and signals from every M'th element of the second array are combined to generate M' group signals. M and M' are different numbers from each other.

Another embodiment relates to an optical positioning device comprising a two-dimensional array of photosensitive elements organized in an M x M' pattern that repeats to form the elements of the array. The circuitry is configured to combine the signals from each element at the same location in the pattern to produce M x M' group signals.

Other embodiments are also disclosed.

Description of drawings

These and various other features and advantages of the present disclosure will be more fully understood from the following detailed description and from the accompanying drawings, which should not be construed to limit the appended claims to the specific implementation shown. By way of example, and for purposes of explanation and understanding only, the accompanying drawings include:

Figures 1A and 1B illustrate the diffraction pattern of light reflected from a smooth surface and the spots in the interference pattern of light reflected from a rough surface, respectively;

Figure 2 shows a functional block diagram of a speckle-based OPD according to an embodiment of the disclosure;

Figure 3 shows a block diagram of an array with interleaved groups of photosensitive elements in accordance with an embodiment of the disclosure;

Figure 4 shows a graph of simulated signals from the array of Figure 3 according to an embodiment of the disclosure;

5 illustrates a block diagram of an array arrangement with multiple rows of interleaved groups of photosensitive elements and the resulting in-phase signals in accordance with an embodiment of the disclosure;

6 illustrates a graph of simulated signals from an array having interleaved groups of photosensitive elements in which signals from every fourth photosensitive element are electrically connected or combined in accordance with an embodiment of the disclosure;

Figure 7 shows an estimated velocity histogram for a detector having 64 photosensitive elements connected in a 4N configuration and operating at 81% of maximum velocity in accordance with an embodiment of the present disclosure;

Figure 8 shows error rate as a function of the number of elements of a detector having photosensitive elements connected in a 4N configuration, according to an embodiment of the present disclosure;

Fig. 9 shows the dependence of the error rate on the signal amplitude according to an embodiment of the present invention;

Figure 10 shows the error rate as a function of the number of elements of a detector having multiple rows of photosensitive elements connected in a 4N configuration, according to an embodiment of the present disclosure;

Figure 11 shows a graph of simulated signals from an array having interleaved groups of photosensitive elements connected in various configurations, according to an embodiment of the disclosure;

Figure 12 shows a block diagram of an array arrangement with photosensitive elements connected in a 5N configuration and original and orthogonal weighting coefficients in accordance with an embodiment of the disclosure;

Figure 13 shows a block diagram of an array arrangement with photosensitive elements connected in a 6N configuration and original and orthogonal weighting coefficients in accordance with an embodiment of the disclosure;

Figure 14 shows a block diagram of an array arrangement with photosensitive elements connected in a 4N configuration and original and orthogonal weighting coefficients in accordance with an embodiment of the disclosure;

FIG. 15 is a block diagram illustrating an arrangement of a multi-row array with photosensitive elements connected in a 6N configuration and a 4N configuration according to an embodiment of the disclosure;

Figure 16 shows a schematic diagram of an embodiment of a circuit using current mirrors to implement 4N/5N/6N weighted groups by reusing the same component output to generate multiple independent signals for motion estimation, according to an embodiment of the disclosure;

Figure 17 illustrates a multi-row array arrangement having two rows connected end-to-end, rather than on top of each other, in accordance with an embodiment of the disclosure; and

Figure 18 illustrates the arrangement of photosensitive elements in a two-dimensional array according to an embodiment of the disclosure.

Detailed ways

Problems with Existing Optical Positioning Devices

One problem with existing spot-based OPDs is due to the pitch or distance between adjacent photodiodes, which typically ranges from ten (10) microns to five hundred (500) microns. Spots with a size smaller than this pitch in the imaging plane cannot be detected correctly, thereby limiting the sensitivity and accuracy of OPD. Spots significantly larger than this spacing will produce a greatly reduced signal.

Another problem is that the coherent light source must be properly aligned with the detector in order to produce a speckled surface image. In existing designs, the illuminated portion of the image plane is typically much wider than the detector's field of view to ensure that the photodiode array is fully covered by reflected illumination. However, having a large illuminated area reduces the power intensity of reflected illumination that the photodiode can detect. Therefore, attempts to solve or avoid the misalignment problem in existing spot-based OPDs often result in a loss of available reflected light from the photodiode array, or place higher demands on the illumination power.

Yet another problem with conventional OPDs is the problem of distortion of features on or emanating from the surface due to viewing angles and/or changing distances between imaging optics and features at different points within the field of view. This is particularly a problem for OPDs that use illumination at tangential incidence.

An additional problem with existing speckle-based OPDs arising from image analysis of speckle patterns is the sensitivity of the estimation scheme to statistical fluctuations. Since speckles are generated by phase randomization of scattered coherent light, speckles have a defined size and distribution on average, but speckles may exhibit local patterns inconsistent with this average. As a result, devices may be prone to locally blurred or difficult to interpret data, eg speckle patterns providing less motion-related signals than usual.

Yet another problem with existing speckle-based OPDs involves changes in the speckle pattern, or speckle "boiling". In general, as the surface moves, the speckle pattern from the surface moves with it, in the same direction and at the same speed. But in many optical systems there is an additional change in the phase front away from the surface. For example, if the optical system is not telecentric such that the path length from the surface to the corresponding detector is not uniform across the surface, the spot pattern may change in a somewhat random manner as the surface moves. This distorts the signal used to detect surface movement, reducing the accuracy and sensitivity of the system.

Therefore, there is a need for a highly accurate speckle-based pointing device and method of using the same that is capable of detecting movement with less than 0.5% path error or up and down. The device is expected to have a simple, understandable and uncomplicated design, with relatively low requirements for image processing. It is also desirable for the device to have high optical efficiency, ie, the loss of reflected light available to the photodiode array is minimized. It is also desirable to optimize the sensitivity and precision of the device to the spot size used, as well as the precise maintenance of the spot pattern by the optical system.

OPD embodiments disclosed herein

The present disclosure generally relates to sensors for optical positioning devices (OPDs), and methods of sensing relative movement between the sensor and the surface based on the displacement of a random intensity distribution pattern of light reflected from the surface, called a spot. OPDs include, but are not limited to, optical mice or trackballs for inputting data into personal computers.

Reference in the specification to "one embodiment" or "an embodiment" means that a specific feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment.

In general, a sensor for an OPD includes: an illuminator, which has a light source and illumination optics to illuminate a portion of the surface; a detector, which has multiple photosensitive elements and imaging optics; and signal processing or mixed-signal electronics with The signals from the individual photosensitive elements are combined to produce the output signal of the detector.

In one embodiment, the detector and mixed signal electronics are fabricated using standard CMOS processes and equipment. Preferably, the sensors and methods of the present invention provide an optically efficient detection architecture using structure illumination and telecentric speckle imaging, and a simplified signal processing configuration using a combination of analog and digital electronics. This architecture reduces the amount of power dedicated to signal processing and displacement estimation in the sensor. It has been found that a sensor using speckle detection technology and properly configured in accordance with the present invention can meet or exceed all performance criteria normally expected of an OPD, including maximum displacement velocity, accuracy and % path error rate.

Introduction to Speckle-Based Displacement Sensors

This section discusses applicants' understanding and belief of how a speckle-based displacement sensor works. While these principles of operation are useful for understanding, embodiments of the invention should not be unnecessarily limited by these principles.

Referring to FIG. 1A , laser light at the indicated wavelengths is shown impinging on 102 and reflecting 104 from a smooth reflective surface, where the angle of incidence Θ is equal to the angle of reflection Θ. A diffraction pattern 106 is produced which has a periodicity of λ/2sinθ.

In contrast, referring to FIG. 1B , any general surface with topological irregularities having dimensions larger than the wavelength of light (ie, approximately >1 μm) tends to scatter light 114 in an approximately Lambertian fashion over the full range. If a coherent light source such as a laser is used, the spatially coherent scattered light produces a complex interference pattern 116 when detected by a square-law detector with a finite aperture. This complex interference pattern 116 of bright and dark regions is called a speckle. The exact nature and contrast of the speckle pattern 116 depends on the surface roughness, the wavelength of the light and its degree of spatial coherence, and the light collecting or imaging optics. Although often highly complex, a distinctive feature of the speckle pattern 116 is that a segment of any rough surface is imaged by the optics and can then be used to identify it as its position on the surface is displaced laterally relative to the laser and optics-detector assembly .

The spot is expected to be of all sizes, up to the spatial frequency set by the effective aperture of the optic, conventionally defined by its numerical aperture NA=sin Θ, as shown in Figure 1B. According to Goodman [J.W.Goodman, "Statistical Properties of Laser Speckle Patterns" in "Laser Speckle and Related Phenomena"edited by J.C.Dainty, Topics in Applied Physics volume 9, Springer-Verlag (1984) - see pages 39-40 for details], size statistics Distributions are represented by spot intensity auto-coherence. The "average" spot diameter can be defined as:

$a=\frac{\mathrm{\λ}}{\sin \mathrm{\θ}}=\frac{\mathrm{\λ}}{\mathrm{NA}}$ (Formula 3)

It is interesting to point out that the spatial spectral density of the speckle intensity, according to the Wiener-Khintchine law, is the Fourier transform of the intensity auto-coherence. The smallest possible spot, _amin = λ/2NA, is set by the improbable case that the main contribution comes from the outermost ray 118 of Fig. 1B (i.e. the ray at ±θ), and from the most "inner" The effect of rays is destructive interference. So the cutoff spatial frequency is f _co =1/(λ/2NA) or 2NA/λ.

Note that the numerical aperture can be different for spatial frequencies in the image along one dimension (eg "x") and along its orthogonal dimension ("y"). This may be due to the optical aperture being longer in one dimension than the other (eg, an ellipse rather than a circle), or due to an anamorphic lens. In these cases, the spot pattern 116 would also be anisotropic and the average spot size would differ in two dimensions.

One advantage of a speckle-based laser displacement sensor is that it can operate with illuminating light arriving at a near-normal angle of incidence. Sensors employing imaging optics and incoherent light hitting rough surfaces at tangential incidence angles can also be used for lateral displacement sensing. However, since the tangential angle of incidence of the illumination is used to produce suitably large light-dark shadows of the surface topography in the image, such a system is inherently optically inefficient since a significant fraction of the light is specularly reflected out of the detector, Therefore it has no effect on the image formed. In contrast, spot-based displacement sensors can efficiently utilize a larger fraction of the illuminating light from the laser source, allowing the development of optically efficient displacement sensors.

Disclosed Architecture of Speckle-Based Displacement Sensor

The following detailed description describes the architecture for one such spot-based laser displacement sensor using CMOS photodiodes with analog signal combining circuitry, appropriate digital signal processing circuitry, and a low power light source such as an 850nm vertical cavity Surface emitting lasers (VCSELs). Although certain implementation details are discussed in the following detailed description, those skilled in the art should understand that different light sources, detectors or photosensitive elements, and/or Different circuits for combining signals can also be used.

Referring now to FIGS. 2 and 3, a speckle-based mouse according to an embodiment of the present invention will be described.

FIG. 2 shows a functional diagram of a blob-based system 200 according to an embodiment of the invention. System 200 includes laser source 202 , illumination optics 204 , imaging optics 208 , at least two sets of multiple CMOS photodiode arrays 210 , front-end electronics 212 , signal processing circuitry 214 , and interface circuitry 216 . Photodiode array 210 may be configured to provide displacement measurements along two orthogonal axes x and y. The groups of photodiodes in each array may be combined using passive electronic components in front-end electronics 212 to generate a group signal. The group signals may then be algebraically combined by signal processing circuitry 214 to produce an (x,y) signal providing information on the magnitude and direction of OPD displacement in the x and y directions. The (x,y) signal may be converted by interface circuit 218 into x,y data 220, which may be output by the OPD. Sensors using this detection technique can have an array of interleaved groups of linear photodiodes, known as a "differential comb array."

Figure 3 shows a general configuration (along one axis) of such a photodiode array 302, where the surface 304 is illuminated by a coherent light source, such as a vertical cavity surface emitting laser (VCSEL) 306 and illumination optics 308, and where the array 302 is staggered The combination of groups acts as a periodic filter on the spatial frequency of the bright-dark signal generated by the speckle image.

Spots from the rough surface 304 are imaged onto the detector plane with imaging optics 310 . Preferably, imaging optics 310 are telecentric for best performance.

In one embodiment, comb array detection is performed in two separate orthogonal arrays to obtain displacement estimates in x and y. A small such array 302 is shown in FIG. 3 .

Each array in the detector includes N groups of photodiodes, each group has M photodiodes (PDs), arranged to form an MN linear array. In the embodiment shown in FIG. 3, each group includes four photodiodes (4PD), referred to as 1,2,3,4. The PD1 in each group is electrically connected (wired and) to form a group, and the same is true for PD2, PD3 and PD4, resulting in four signal lines coming out of the array. Their corresponding currents or signals are I ₁ , I ₂ , I ₃ and I ₄ . These signals (I ₁ , I ₂ , I ₃ and I ₄ ) may be referred to as group signals. By using the differential analog circuit 312 to generate the in-phase differential current signal 314(I ₁₃ )=I ₁ −I ₃ , and using the differential analog circuit 316 to generate the quadrature differential current signal 318(I ₂₄ )=I ₂ −I ₄ , it can be achieved Background suppression (and signal emphasis). These in-phase and quadrature signals may be referred to as line signals. The direction of movement can be detected by comparing the phases of I ₁₃ and I ₂₄ .

One difficulty with comb detectors using 4N detection as shown in FIG. 3 is that they can have unacceptably large error rates unless they have very large arrays, such as more than a few hundred detectors in array 102 or photodiodes. These errors arise when the oscillating signal is weak due to an effective balance between the intensity of light falling on different parts of the array. In the simulation in FIG. 4, for example, in and around frame 65, the amplitude of the oscillation signal is relatively small. Referring to Figure 4, the in-phase (original) and quadrature signals are shown. Frame numbers are shown along the horizontal axis.

multi-row detector array

One solution to this fundamental source of noise is to have rows of these detectors or photosensitive elements arranged in sets or together. A detector with two nested rows 502-1 and 502-2 is shown schematically in FIG. The resulting oscillating in-phase signals 504-1 and 504-2 from these two rows are also shown. In this type of detector, when one row produces a weak signal, velocity can be measured from the signal from the other row. For example, around frame 2400, the in-phase signal 504-1 has a relatively small amplitude, but the second in-phase signal 504-2 has a relatively large amplitude. As shown below, when the oscillation amplitude is large, the error rate is small. Therefore, the "right" row (ie, the row with relatively larger amplitude oscillations) can be selected and evaluated with low error.

simulation method

To demonstrate the efficacy of the Figure 5 configuration, a speckle pattern was generated on a square grid with random and independent intensity values in each square. The spot size, or grid spacing, was set at 20 microns. Another grid representing the detector array is generated, which has variable dimensions and is scanned at a constant speed over the spot pattern. The instantaneous intensity at each detector or photosensitive element is summed with other photocurrents in the same group to determine the signal. The following simulations use a "4N" detector scheme with constant horizontal detector or photosensitive element spacing.

Error rate calculation

An example output of these simulations is shown in Figure 6, which shows simulated in-phase (raw) signal 602-1 and quadrature signal 602-2 from a 4N comb detector. Also shown is the magnitude (length) 604 and phase (angle) 606 of the vector defined by these two signals. In this exemplary simulation, each array includes 84 detectors or photosensitive elements, operating at 5% of the maximum speed.

The horizontal axis on these graphs shows frame counts; in this case 4000 individual measurements (frames) were used. The lower two curves are the in-phase 602-1 and quadrature 602-2 signals (group 1 minus group 3, and group 2 minus group 4, respectively). From these two curves, the signal length 604 and angle 606 can be determined, as shown in the upper two curves. Note that the in-phase 602-1 and quadrature 602-2 signals are very similar since they depend on the same part of the speckle pattern.

This data can be used to calculate speed. In this example, we use a simple zero-crossing algorithm for velocity calculations. At each frame, calculate the number of frames τ between the previous two positive going zero crossings. A positive going zero crossing is a zero crossing where the slope of the line is positive to move the signal from negative to positive. In this case, τ represents an estimate of the number of frames required to advance 20 micrometers (μm). Consider the frame rate (frames per unit of time) as f and the detector pitch (distance from the start of a group of elements to the next group of elements) as p. Then the estimated velocity (velocity) v is:

v=f*p/τ (Formula 4)

The maximum velocity v _max is half the Nyquist velocity. A histogram of the results is shown in FIG. 7 .

Referring to Figure 7, the histogram shows the estimated speed for a 64 photosensitive element detector, 4N detector operating at 81% of the maximum speed. The vertical line 701 at frame 4.938 represents the actual speed estimated from the data. Different point markers in the histogram are used for different selections of data sets: the first marker 702 indicates the number of occurrences when all frames are included; the second marker 704 indicates when those frames in the lower 17% of the magnitude distribution are not included the number of occurrences; the third mark 706 represents the number of occurrences when those frames in the amplitude distribution below 33% are not included; the fourth mark 708 represents the number of occurrences when those frames in the amplitude distribution below 50% are not included; and The five marks 710 indicate the number of occurrences when those frames in the lower 67% of the amplitude distribution are not included.

The points of the first marker 702, containing all the data, show a distribution with a strong peak at frame 5 and a rapid decrease to both sides. The vertical line 701 at frame 4.938, which we call the "true value", is the estimated actual speed. On each side of this line (i.e. at frames 4 and 5) there are two relatively strongest peaks in the data.

For the purposes of this simulation, we count any point that falls outside of these two strongest peaks as an error. In other words, estimates that are more than one frame away from the "truth" are defined as "errors". This is a fairly strict definition of an error, since it often occurs in subsequent cycles. If the actual velocity is at a frame close to an integer, there will be a large fraction of the error that is only slightly more than one frame from the "true value". For example, the estimated "true value" of 4.938 frames at 6 frames in Figure 7 is only slightly more than one frame. Under this rather strict definition, these points in 6 frames would all be considered "errors".

Figure 8 shows the error rate as a function of the number of elements in a 4N detector. Referring to Figure 8, it can be seen that the error rate decreases with increasing number of detectors or photosensitive elements, as expected from previous work.

For these measurements, error rates were calculated and averaged over seven (7) different speeds.

Relationship to vector length

Errors are concentrated in those frames with weak signals. The data in Figure 7 also shows a histogram of the data after selection of the vector magnitude. For example, the points marked 706 are velocity estimates for only those frames with vector lengths in the upper two-thirds of the profile (ie, excluding the lower 33% based on signal magnitude or signal vector length). So the data excludes those frames where the signal is weak and expected to be prone to errors. As expected, when smaller signal amplitudes are excluded, the distribution of frame numbers between zero crossings is narrower, resulting in a significant improvement in the calculated error rate.

The improvement in error rate by excluding smaller signal amplitudes is shown in Figure 9. Figure 9 shows the error rate versus signal amplitude. More specifically, the error rate is shown as a function of the minimum percentile of the signal vector length used. Referring to Figure 9, it can be seen that the upper two-thirds of the vector length distribution (represented by data point 902) has an error rate that is only one-third that of all frames (represented by data point 904): i.e. 4.8% versus 14.1 %. Using only the upper third (represented by data point 906) further reduces the error rate to 1.2%.

Therefore, one approach to row selection from among the rows of the detector is to select the row with the highest signal magnitude, in terms of error rate improvement when smaller signal magnitudes are excluded. For example, in the case of Figure 5 with two nested rows, the signal from the second row 504-2 is selected for frame 2400 because there is a larger amplitude at that point, while the signal from the first row 504-1 was chosen for frame 3200 because there is a larger margin at that point. Of course, this selection scheme can be applied to more than two rows. Furthermore, instead of using signal amplitude (AC strength) as a measure of line signal quality, other quality measures or indicators may be used.

Selecting the line signal from the line with the highest line signal quality is a scheme to avoid or counteract speckle fading using signals from multiple lines. Additionally, there are various other alternatives for accomplishing the same or a similar purpose.

An alternative is to weight the line signals from different rows by their amplitude (or other quality measure) and then eg average the weighted signals again. In one embodiment, rather than simply averaging the weighted signals, the group of weighted signals is more optimally processed by an algorithm using recursive filtering techniques. A notable example of a linear recursive filtering technique uses a Kalman filter. [See R.E. Kalman, "A New Approach to Linear Filtering and Prediction Problems," Trans. ASME, Journal of Basic Engineering, Volume 82 (Series D), pp. 35-45 (1960).] The extended Kalman filter can be used for nonlinear estimation Algorithms (eg in the case of sinusoidal signals from a comb detector arrangement). Signal properties and measurement models for a speckle-based optical mouse show that a recursive digital signal processing algorithm is well suited to the weighted signals produced by the speckle mouse front-end detector and electronic circuitry.

Simulation of multi-line arrangement

Two-row and three-row detectors are simulated using the same technique. Each row is illuminated by a separate part of the speckle pattern. The results of the error rate are shown in Fig.10.

FIG. 10 shows the error rates for motion detectors having three (3) rows of 4N detectors 1002 , having two (2) rows of 4N detectors 1004 , and one (1) row of 4N detectors 1006 . Trendlines are also shown for the 3-row data 1012 , the 2-row data 1014 , and the 1-row data 1016 . These error rates were calculated by averaging the results at three (3) different speeds over five thousand (5000) frames. Multiple points on the graph represent different simulations: we use four different rows for 1-row measurements; three different combinations of two rows for 2-row measurements; and two different combinations of three rows for 3-row measurements . To ensure unbiased comparisons, two-row and three-row data were formed by combining the original four rows.

The simulations show, for example, that a single row of 32 elements has an error rate of slightly greater than 20%. Combining two such rows (for a total element count of 64) reduces the error rate to about 13%. This is slightly lower than the result for a single row of 64 elements. Combining three such rows (total element count of 96) gives an error rate of about 8%, dropping to less than 1/2 the error rate of a single row.

For higher component counts, the benefit of increasing the number of rows is greater. Combining three rows of 128 elements in each row (the total number of elements is 384) reduces the error rate from 10% (128 elements in a single row) to 1.5% (combination of three such rows), down to less than 1/ of the error rate of a single row 6.

path error

From this error rate, we can calculate the path error as follows. In traversing a path that is M count long, the total number of errors is ME. Here E is the error rate discussed and calculated above. Errors appear as extra and missing counts as the surface moves. For measurements over longer distances, these errors tend to cancel out, and the average net error only increases with the square root of the total number of errors. The measured count can differ from the expected count by a positive or negative amount, but on average it has an absolute value equal to the square root of the number of errors. We define path errors as:

$\mathrm{path}\_\mathrm{error}=|1-\frac{\mathrm{Measured}\_\mathrm{counts}}{\mathrm{Expected}\_\mathrm{counts}}|$ (Formula 5)

计数结束。所以，在所测量的计数高于预期的计数时， $\mathrm{Measured}\_\mathrm{counts}=M+\sqrt{\mathrm{ME}},$ 且路径差错为：When traversing a path with a length of M counts, the mouse will generate ME errors on average, and the

The count is over. Therefore, when the measured count is higher than the expected count, $\mathrm{Measured}\_\mathrm{counts}=m+\sqrt{\mathrm{ME}},$ And the path error is:

$\mathrm{path}\_\mathrm{error}=|1-\frac{m+\sqrt{\mathrm{ME}}}{m}|=\sqrt{\frac{\mathrm{E.}}{m}}$ (Formula 6)

This is only a rough representation of the mean path error, in a more precise calculation it has a distribution centered around zero with a standard deviation of

To apply this formula to the above results, we assume a resolution of 847 dots per inch (dpi) (ie 847 frames or samples per inch) and a distance of 2 centimeters (cm) of travel. This gives 667 frames per measurement (ie 667 frames at 2 cm advance), so M=667. For three rows of 128 detectors or photosensitive elements each, we obtain an error rate E of 1.5%, so according to Equation 6, the path error is 0.5%. At longer distances, the path error improves significantly.

Detection using detectors or sets of photosensitive elements

Another solution to the noise problem of comb detectors using 4N detection is to provide a detector whose array includes one or more rows of groups (N) of interleaved photosensitive element groups, each group having a number of consecutive photosensitive elements (M), where M is not equal to four (4). In other words, M is a number from the group consisting of 3, 5, 6, 7, 8, 9, 10, etc. Specifically, every third, every fifth, every sixth or every Mth detector or photosensitive element is combined to generate an independent signal for estimating motion.

Figure 11 shows the raw and quadrature signals for combining every third 1102, every fourth 1104, every fifth 1108, and every sixth 1110 detector or photosensitive element, and operating at the same detection intensity . The signal shown in Figure 11 is a simulated signal from an array with interleaved groups of photosensitive elements or detectors, where the raw detections from every third, fourth, fifth and sixth detector or photosensitive element are combined. Referring to Figure 11, the original and quadrature signals are shown, with frame numbers given along the horizontal axis. As can be seen from Figure 11, when one subgroup of detectors or photosensitive elements produces a weak signal, the other subgroup can be used to measure velocity. As mentioned above, when the oscillation amplitude is large, the error rate is small. Therefore, the "right" (larger amplitude) signal can be selected and a low error estimate can be made.

The above example includes one hundred and twenty (120) detectors or photosensitive elements operating at about 72% of the maximum rated speed. The horizontal axis on the graph in FIG. 11 shows the frame count. Note that the original or in-phase signal and the quadrature signal are very similar in that they depend on, or result from, the same speckle pattern.

As mentioned before, this data can be used to calculate speed. In this case we use a simple zero-crossing algorithm. At each frame, calculate the number of frames τ between the previous two positive going zero crossings. This represents an estimate of the number of frames required to advance 20 microns. Consider the frame rate (number of frames per unit of time) to be f, and the detector pitch (distance from the start of a group of elements to the next group of elements) to be p. Then the estimated velocity v is:

v＝f*p/τ

This velocity is a component of the total velocity along the long axis of the detector array.

For configurations other than 4N, groups of detectors or photosensitive elements are weighted and combined in order to generate a velocity-dependent signal. One embodiment of a suitable weighting factor is given by the following formula:

$S1\left(i\right)=\cos (\frac{2^{*}p{i}^{*}i}{m}+\mathrm{phi})$ (Formula 1)

as well as

$S2\left(i\right)=\sin (\frac{2^{*}p{i}^{*}i}{m}+\mathrm{phi})$ (Formula 2)

where i spans all photosensitive elements in the group from 0 to M-1. Here phi is the phase shift common to all weighting coefficients.

The in-phase weighted sum of the output signals (i.e., the in-phase signal) is given by:

$\mathrm{InphaseSum}\left(t\right)=\underset{i=0}{\overset{m-1}{\mathrm{\Σ}}}S1{\left(i\right)}^{*}\mathrm{DetectorOutput}(i,t)$ (Formula 7)

And the quadrature weighted sum of the output signal (i.e. the quadrature signal) is given by:

$\mathrm{Quadrature\; Sum}\left(t\right)=\underset{i=0}{\overset{m-1}{\mathrm{\Σ}}}S2{\left(i\right)}^{*}\mathrm{DetectorOutput}(i,t)$ (Formula 8)

These coefficients are shown in Figure 12 for a 5-element group, ie for a 5N configuration. In this example, five lines "sum" (1202-1, 1202-2, 1202-3, 1202-4, 1202-5) are formed. The original signal is the sum of each line sum multiplied by its original weight, where the original weight of each line sum is given by column S1 in Fig. 12. Likewise, the quadrature signal is the sum of each line sum multiplied by its quadrature weight, where the quadrature weight of each line sum is given by column S2 in Fig. 12.

The weighting coefficients for an array with photosensitive elements connected in a 6N configuration are shown in FIG. 13 . The original weighting coefficients corresponding to the six-line sum are given under the S1 column, and the orthogonal weighting coefficients corresponding to the six-line sum are given under the S2 column.

The weighting coefficients for an array with photosensitive elements connected in a 4N configuration are shown in FIG. 14 . The original weighting coefficients corresponding to the four-line sum are given under the S1 column, and the orthogonal weighting coefficients corresponding to the four-line sum are given under the S2 column. For a 4N comb, the weighting coefficients are all 0 or +/-1, and the system can be simplified as a differential amplifier, as shown in and described in conjunction with FIG. 3 .

In another aspect, the present disclosure is directed to sensors having detectors with two or more differently grouped photosensitive elements. Such an embodiment with multiple groupings of elements allows multiple independent signals to be generated for motion estimation.

For example, if combs with different values of M are combined in the same sensor (e.g. 4N and 6N), and the width of the photosensitive element is kept constant, we can see from the arrangement shown in Figure 15 that there are different but parallel arrays get good performance. 15 is a block diagram of an arrangement of a two-row array with photosensitive elements connected in a 6N configuration 1502 and a 4N configuration 1504 in accordance with an embodiment of the invention. In this case, two different spot patterns are measured, one for each row.

Alternatively, we can use the same array and the same part of the speckle pattern. This is the situation modeled in Figure 11 above. The advantage of this approach is saving photodiode space, as well as the leakage current associated with each photodiode. It also conserves photons, since smaller areas on the silicon need to be illuminated with a speckle pattern.

A circuit implementation for wiring individual photodiode elements having multiple values of M is shown in FIG. 16 . Figure 16 is a schematic diagram in which current mirrors are used to implement 4N, 5N, and 6N weighted groups in a manner that reuses the same component outputs in accordance with an embodiment of the present invention. The circuit 1600 of FIG. 16 generates multiple independent signals for motion estimation, each for a different M configuration. In this example, current mirrors 1604 are used to double the output current of each detector or photosensitive element 1602 . The currents are summed by connecting these outputs together using a wiring structure 1606 ordered in different M configurations. For multiple values of M, these wiring structures 1606 add together every Mth output current. The weighted magnitude is applied by the current reduction element 1608 . For each in-phase and quadrature output, another wiring structure 1610 sums together the currents for the positive weights and sums together the currents from the negative weights, respectively. Finally, for each in-phase and quadrature output, a differential circuit 1612 receives the respective currents for the positive and negative weights and produces an output signal.

In the specific example shown in Figure 16, for M=4, 5 and 6, separate in-phase and quadrature outputs are produced. In other implementations, in-phase and quadrature outputs can be generated for other values of M. Also, the in-phase and quadrature outputs can be generated for more (or fewer) values of M, not necessarily exactly 3 values of M as in the specific example in FIG. 16 .

In an alternative circuit implementation, each detector or photosensitive element can feed multiple current mirrors with different gains, so that the same detector or photosensitive element can be used for different detector periods (M values) for different, Independent in-phase and quadrature sums work.

In another alternative circuit implementation, the detector values may be sampled individually using an analog-to-digital converter (ADC) circuit, or the detector values may be multiplexed and sampled sequentially, and the digitized values may then be processed to produce independent sum. In yet another circuit implementation, the analog sum of the detector outputs can be processed by a shared time-multiplexed or multiple simultaneous ADC circuits. There are many circuit implementations to accomplish this task, with different implementations weighing various factors such as circuit complexity, power consumption, and/or noise figure.

The embodiments shown in Figures 5 and 15 show a one-dimensional array of rows. The rows are joined along their short axis - one on top of the other. Alternatively, it is also useful to have two rows joined along the long axis, as shown in FIG. 17 .

In FIG. 17 , a single one-dimensional array is divided into two parts, left 1702 and right 1704 . Each side can be configured in a comb arrangement with the same M value. In the specific implementation scheme of FIG. 17, M=5. Other implementations may use other values of M. The left side 1702 produces a set of signals 1706 and the right side 1704 produces a second set of signals 1708 . These two sets of signals are optionally combined into a third set of signals 1710 . Thus, there are three groups of signals from which to choose, based on signal magnitude or other mechanisms as described above. This arrangement has the advantage that the combined set of signals 1710 benefits from an effectively longer array, which should have excellent noise properties.

The above detailed embodiments show detectors or photosensitive elements oriented along a single axis - ie in a one dimensional array, although there may be several rows. In another embodiment, the detectors or photosensitive elements are a two-dimensional array, eg, as shown in FIG. 18 .

In Figure 18, an example two-dimensional (2D) array of 21x9 elements is arranged in groups of 9 elements (in a 3x3 matrix). Elements (shown with the same color) at a given position in a group are grouped into groups by common wires. With this configuration, motion information in x and y can be collected by the same set of detectors or photosensitive elements. While each group is a 3x3 matrix in the example 2D array of Figure 18, other implementations may have groups of other dimensions. A set may have a different number of elements along the horizontal dimension (x) 1802 than the number of elements along the vertical dimension (y) 1804 . Furthermore, while the photosensitive elements shown in FIG. 18 are all equally sized and rectangular, alternative implementations may use photosensitive elements of different sizes and/or shapes that are not rectangular.

The foregoing descriptions of specific embodiments and examples of this invention have been presented for purposes of illustration and description, and while the invention has been illustrated by certain foregoing examples, it should not be construed as limiting thereto. The descriptions and illustrations are not intended to be exhaustive or to limit the invention to the precise form disclosed, and many modifications, improvements and variations are possible in light of the above teaching and are within the scope of the invention. The scope of the present invention shall include the general field as disclosed herein and covered by the claims appended hereto and their equivalents.

Claims (20)

1. optical displacement sensor is used for by determining to come relatively moving between sense data input media and the described surface in the displacement of the successive frame optical characteristics on surface, and described sensor comprises:

Detecting device comprises a plurality of light activated elements that are organized into first and second arrays;

First circuit is configured to make up the signal from every M the element of first array, to produce the M group signals; And

Second circuit is configured to the signal of combination from the individual element of the every M ' of second array, with generation M ' group signals,

Wherein M and M ' are the numbers that differs from one another.

2. optical displacement sensor as claimed in claim 1, wherein first and second arrays comprise at least one shared light activated element.

3. optical displacement sensor as claimed in claim 2 also comprises:

Current mirroring circuit is configured to make the photocurrent from described shared light activated element to double.

4. optical displacement sensor as claimed in claim 2, wherein first and second arrays are shared all described light activated elements basically.

5. optical displacement sensor as claimed in claim 4 also comprises:

Current mirroring circuit is configured to make from the photocurrent of all described light activated elements basically double.

6. optical displacement sensor as claimed in claim 1, wherein each described array comprises the linear comb arrays (LCA) that is parallel to the axle arrangement.

7. optical displacement sensor as claimed in claim 6 also comprises:

Be configured to from described M group signals, produce the circuit of first pair of line signal; And

Be configured to from described M ' group signals, produce the circuit of second pair of line signal.

8. optical displacement sensor as claimed in claim 7, wherein when described data input device when the direction that is not orthogonal to described axle moves, described line signal vibrates in form.

9. optical displacement sensor as claimed in claim 8 also comprises:

Be configured to use the circuit of the speed component v that following formula moves along described axle according to the described data input device of line calculated signals:

v＝f*p/τ

F is a frame rate, and p is a detector spacings, and τ is the time between preceding two zero crossings of same direction trend.

10. optical displacement sensor as claimed in claim 7, wherein said line signal generating circuit comprises weighting circuit, with the described group signals of weighting.

11. optical displacement sensor as claimed in claim 10, wherein said M group signals is respectively by a M homophase weighting coefficient and the weighting of a M quadrature weighting coefficient, and wherein said M ' group signals is respectively by a M ' homophase weighting coefficient and the weighting of a M ' quadrature weighting coefficient.

12. optical displacement sensor as claimed in claim 11, wherein said homophase weighting coefficient (S1) and described quadrature weighting coefficient (S2) use following formula to calculate:

$S1=\cos \frac{2\mathrm{\πj}}{N};$

And

$S2$ $=$ $\frac{2\mathrm{\πj}}{N},$

J is from 0 to M-1 in the formula, and N=M, is used for described M group signals, is used for producing first pair of line signal, and

J is from 0 to M '-1 in the formula, and N=M ', is used for described M ' group signals, is used for producing second pair of line signal.

13. optical displacement sensor as claimed in claim 1 also comprises:

Be configured to circuit that the line quality of signals that produces and another line quality of signals that produces are compared from described M ' group signals from described M group signals, and

Be configured to the circuit of relatively selecting better quality line signal based on described.

14. optical displacement sensor as claimed in claim 1 also comprises:

Be configured to based on signal quality characteristics weighting line signal and make up the circuit of the line signal of institute's weighting.

15. a method of using optical displacement sensor sense data input media to move from the teeth outwards, the detecting device that described optical displacement sensor has comprises a plurality of light activated elements that are organized into first and second arrays, and described method comprises:

On described a plurality of light activated elements, receive the intensity pattern that light produced from the described surface reflection of part;

Combination is from the signal of every M element of first array, to produce the M group signals; And

Combination is from the signal of the individual element of every M ' of second array, producing M ' group signals,

Wherein M and M ' are the numbers that differs from one another.

16. method as claimed in claim 15, wherein first and second arrays are shared light activated element.

17. method as claimed in claim 15 also comprises:

From described M group signals, produce first pair of oscillating line signal; And

From described M ' group signals, produce second pair of oscillating line signal.

18. method as claimed in claim 17 also comprises:

The line quality of signals that produces from described M group signals and another line quality of signals that produces from described M ' group signals are compared; And

Based on the described line signal of relatively selecting better quality.

19. method as claimed in claim 17 also comprises:

Based on signal quality characteristics weighting line signal; And

The line signal of combination institute weighting.

20. an optical positioning device, described equipment comprises:

The light activated element of two-dimensional array is organized into the M * M ' pattern that is repeated with the element that forms described array; And

Be configured to make up signal from each element on the same position in the described pattern so that produce the circuit of M * M ' group signals.

Images