JPH06509449A - smart tracking system - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] smart tracking system (Related application) This application is filed in pending U.S. patent application Ser. No. 87, current U.S. Pat. No. 4,980, the subject matter of which is incorporated herein by reference. ．． No. 871 rUILrasonic Tracking SystemJ .

(Technical field) Regarding measurement and tracking systems, especially when cameras track moving objects. It concerns a system in which the method of movement depends on the context and history of the subject's movement.

(Background technology) Automatic tracking system as disclosed in U.S. Pat. No. 4,980,871 There are many aesthetic factors to consider when directing a video camera for a human viewer. be. In particular, centering the subject within the camera frame and against a stable background. Factors such as the smoothness of the movement of the camera and the continuity of the direction of movement of the camera These are all the elements that bring the aesthetic pleasure of video images to the viewer. On the contrary, suddenly Harsh or continuous camera movements and inconsistent camera tracking speeds Video images characterized by a tendency to undermine ICC youth visual satisfaction standards has.

11 Automated tracking systems such as g2 A two-step process is used to remove errors between the la directions. First, this tiger A tracking system provides a link between the tracked subject and the direction of the tracking device/camera. Measure the angular error. Next, the error information is used to operate the tracking device/camera. produces a correction signal that controls the motion. The new camera position is then adjusted for future angular errors. Used during calculations. The resulting tracking is 2 tends to be very unpleasant for the winners as it is also very inaccurate. It is mechanical. For example, the camera position for each movement of the subject A quick readjustment can cause rapid eye fatigue for the viewer.

Furthermore, typical automated token sequences are for all subjects and situations, regardless of whether they are participating in the aesthetic performance of use the same tracking method and style. Such tracking systems or change between such styles. The lack of such systems significantly limits the use of such systems.

Unlike automated tracking systems, human photographers are not perfect. Point the camera at the subject with precision. The cameraman also follows the conditions in which the subject moves. to adjust the tracking style and the immediate history of the subject's movement. As a result The resulting video provides the viewer with a clear and precise mechanical tracking of the automated system. Contains minimal tracking error and is more visually pleasing than the standard version.

It is therefore an object of the present invention to determine the location of the subject and the position at which the tracking device/camera is directed. By selectively tolerating angular errors between directions, visual and systems that allow cameras to track moving subjects in an aesthetically pleasing way. On offer.

Another object of the invention is to move the moving subject in a particular style appropriate to the situation in which the subject is moving. The aim is to provide a system that allows cameras to track people.

Another object of the present invention is to determine the motion of the subject based on the past history of the subject's motion. The purpose of the present invention is to provide a system that allows a camera to track a moving subject in a style that allows the camera to track a moving subject.

(Summary of the invention) The system for allowing the device to track the movement of the subject is based on a signal software. a control device that receives signals from the source and generates control signals from the source. include. A storage device stores a plurality of sets of data values, each set being associated with a particular mode of operation. represents the operating parameters of the system in the mode. one of the plurality of sets of data values A device is provided for selecting one. The processor controls according to the selected data value. The control signal is processed and the processed control signal is outputted. The drive unit receives the processed control In response to control signals, the device moves around its axis and remains directed toward the subject. do.

In one embodiment, the apparatus for selecting one of the plurality of sets of data values is configured to select one of the plurality of sets of data values. Contains user-defined signals that indicate data values and their corresponding operating parameters. nothing. In a second embodiment, the device for selecting the set of data values processes a control signal. , determine whether the selected set of data values is appropriate for the moving subject. It is a function of the processor. If not, the processor returns the new data value , and thereby change the operating mode of the system.

These and other features, objects and advantages of the invention will be explained below with reference to the drawings. It will be better understood by reading the detailed description.

(Brief explanation of the drawing) FIG. 1 is a schematic diagram of a drive arrangement of a tracking system suitable for use according to the invention. , Figure 2 shows a transmitter of a tracking system suitable for use according to the present invention. A schematic diagram of a subject such as an athlete, FIG. 3 is a functional block diagram of the drive device according to the present invention, and FIG. 4 is a conventional automatic truck diagram. Flowchart of the error correction process of the locking system; FIG. 5 shows the error correction performed by the drive device of FIG. 3 according to a first embodiment of the invention. process flowchart, FIG. 6A is a flowchart of the algorithm steps for calculating the immobile band for the process of FIG. Chart, FIG. 6B is a flowchart of the algorithm steps for shifting the stationary band of the process of FIG. -chart, Figure 60 shows a flowchart of algorithm steps for continuity control of the process of Figure 5. -chart, FIG. 6D is a flow of algorithm steps for motion side control of the process of FIG. chart, FIG. 7 shows the error correction performed by the drive device of FIG. 3 according to a second embodiment of the invention. process flowchart, and Figure 8 shows the algorithm steps for automatic situation determination of the process of Figure 7. It is a flowchart.

(Example) The present invention provides an active tracking system that includes, in one embodiment, two units. The company provides a smart tracking system that is based on the system. The first unit is , a small battery-powered omnidirectional ultrasound transmitter carried by the tracked subject. It is a trustworthy machine. This transmitter sends short bursts of sound at frequencies beyond the range of human hearing. occurs. The second unit can tilt and pan the attached camera. Includes a possible motor-driven camera drive. The present invention also provides for Achieved by a passive tracking system without the need for an integrated transmitter. You should understand that.

In one embodiment, a battery-powered ultrasound receiver is attached to the camera drive. I get kicked. The receiver is tuned to the same frequency as the transmitter worn by the subject.

The receiver receives the ultrasound signals generated by the transmitter and received by the microphone. Three ultrasonic microphones are used with detection and amplification electronics. My Crophon is an imaginary rectangular quadrangle centered on the axis of the camera lens and perpendicular to it. Placed in three of the corners. The signal emitted by the subject causes the camera to point straight at the subject. It will reach all three microphones at the same time. The receiver is my Measures the relative arrival time of the signal received by the Crophon, and determines whether the camera is the main A servo system that determines the direction and degree of straightness and drives the motor. the camera (and receiver) to face the subject. this The reorientation, or tracking, is done in a manner that is visually pleasing to a human observer. and can be adapted to various situations.

In FIG. 1, a television or movie camera or other device 10 (hereinafter referred to as In the following text, without loss of generality, we will refer to it as a camera), which is designated overall by 12. It is attached to a platform or support that is part of the motor drive and The moving device is further mounted on a tripod 14 or other support. The drive device 12 is a camera around the horizontal axis 16 from 30° below the horizontal to about 30° above the horizontal. It can be tilted up and down. This drive also moves the camera along the vertical axis 18. You can pan left and right repeatedly around you in a 360° circle. drive device to the ultrasound signals received via microphones 2OA, 20B and 20C. In response, controlling a DC motor (not shown) to move the camera support. .

The subject 22 as shown in FIG. An ultrasonic transmitter 24 is provided. For example, the transmitting unit is placed around the subject's waist. It includes an elastic belt 26 with a buckle for attachment. This belt has belt 2 6 or more transducers connected by flexible cables embedded in the -sa device i! 2 (only one transducer device 28 is shown in the figure) is installed. I'm being kicked.

The drives 12 each include two DC motors 42,44. well-known machine According to the structure, the motor 42 tilts the camera support part up and down, and the motor 44 tilts the camera support part up and down. Pan the support from side to side. The camera is fixed relative to the support, so the same movement receive. The motor is mounted on a circuit board (not shown) inside the drive housing. It is driven by an electric circuit. , this circuit connects the battery (also not shown) to Powered by

Three probes for each of the three ultrasonic transducers 20A, 20B and 20C. Output circuits 30A, 30B, 30C are shown in the block diagram of FIG. Also, 2 pieces There are also two motor drive circuits 32.34, one for each of the DC motors 42.44. shown. A control circuit 36 and a digital processor 46 motorize the detector. Connected to the drive circuit. Transducers 20A120B and 20C are each receives an ultrasound burst emitted by a transmitter and sends a corresponding electrical signal to it. to the associated detection circuit. The detection circuit detects the presence of an ultrasound burst and produces a shaped pulse whose leading edge coincides with the beginning of the detected burst. these Pulses 50A, 50rl, 50C are sent to the controller 1-roller 36, and these three pulses are sent to the controller 1-roller 36. A digital error register (not shown) is measured by measuring the delay between pulses from the detector. 2 transmission signals VERR and HERR are loaded into . These 4-bit signals represent the angular error along the vertical and horizontal axes between the camera direction and the subject, respectively. vinegar.

VERR and HERR are updated each time an ultrasound burst is received.

The structure and function of the transmitter 24 including the DC motor 42 and 44, and the camera are the main components. To continue, the method by which drive unit 12 tracks the location of transmitter 24 is as follows: Discussed in detail in U.S. Pat. No. 4,980,871, incorporated herein by reference. and will not be repeated for brevity.

Digital processor 46 provides VERR and H on buses 52.54, respectively. Receive ERR signal. Processor 46 may be a microprocessor. To the present invention One such microprocessor suitable for use in Model MC68705C8 available from MoLolola by hoenix. be. Processor 46 determines the VE according to the particular tracking style required. The values of RR and HERR are manipulated in a manner described in more detail below. These processes The detected error signals PVERR and PHERR are respectively The same drive that drives the tilt and pan motors in the correct direction and at the correct speed to by horizontal (i.e., pan) and vertical (i.e., tilt) motor drive circuits 32.34. Converts to DC motor drive voltage.

(Smart Tracking Process/Smart Tracking Pr.

cesses) A typical automated tracking system follows the basic two-step process shown in Figure 4. Remove the error between the tracked subject's position and camera orientation using the process of (.

First, the tracking system, as shown in block 7o in FIG. Measure the angular error between the tracked subject and the tracking device/camera. Next, block As indicated by lock 72, this error information is used to generate a correction signal to control the data speed and movement of the tracking device/camera. Then this truck The new position of the king device/camera is used when calculating future angular errors.

In the present invention, a number of methods are used to adjust the response characteristics of the tracking system. Signal processing technology attributes the resulting video to human viewers. I can do it. In a first embodiment, the error signals VERR and HERR are 2 bits provided via bus 60 indicating the user-supplied 5 Processed by digital processor 46 according to the TYLE signal.

This 5TYLE signal is sent to the digital processor 46 or to a separate semiconductor memory. select a set of parameters stored in the memory. These parameters are selected Express the acceptable system constraints on the style or mode of tracking real-time data about the location of tracked subjects as well as many data variables. Contains the variables that represent it. The system parameters are set by processor 46. is used to perform a series of calculations to adjust the values of the error signals VERR and HERR. . Processed signals PVERR and PHERR are routed to buses 62.64 and 1, respectively. = + controls the motor drive circuits 32 and 34, respectively, which further controls the viewer's The camera 10 is configured such that the resulting motion of the eye approximates as much as possible the comfortable constraints. direct the person towards the subject.

In a second embodiment, the user provides 5 The TYI, E signals are tracked as described by the most recent values of some variables. Based on the past history of the subject's movement, the digital processor The signal generated by the sensor 46 is replaced.

FIGS. 5 and 7 illustrate digital processing according to the first and second embodiments, respectively. Process executed by processor 46 to process signals VERR and HERR. It shows the This process adjusts the response characteristics of the tracking system. This gives it attributes that result in a video that is pleasing to human viewing.

What is represented by block 78.80.82.84.86.88.90.92 Each of these algorithmic processes is discussed in detail below.

In the teachings of the present invention, the processing of the vertical error signal VERR includes the processing of the horizontal error signal VERR. The processing is substantially the same as the processing of the difference signal HERR. Therefore, the concept of the present invention Only the processing of No. HE RR will be explained.

According to a first embodiment of the invention, the user/operator of the tracking system selects a predetermined style or mode in which the subject will be tracked. This selection is As indicated by lock 78, this occurs simultaneously with powering up the system. The user , the selected mode is digitally indicated by the 2-bit 5TYLE signal applied to line 60. 46. The processor 46 is initialized upon power-up. Select a predefined set of parameters representing the tracking mode or style of can be arbitrarily programmed to Simultaneously with the selection of a predetermined mode, the processor 4 6 all relevant variables as appropriate in relation to the particular tracking style selected. set to a value.

The location of the tracked subject and the tracking device as shown by block 80 of FIG. @/A professional that measures the horizontal and vertical angular error between the directions the camera is pointed at. The process was previously described and the error signals VERR and HERR perform these functions. with points provided to digital processor 46. Similarly, block 90 of FIG. 4-bit off to control motors 42 and 44 respectively as shown by Currently granted U.S. patent applications for the conversion of set binary signals to DC voltages 07/396,987 and will not be described further. Block The algorithm steps indicated by 82.84.86.88 are 5TYL E signals are implemented by digital processor 46 and are described in detail below. described.

(Fixed belt controller: Dcadzonc Ctontroller) Human model digital processor 46 to produce a tracking style that is more pleasing to the winner. The first process executed by is illustrated by processing block 82 in FIG. It includes a dead band controller implemented with the algorithm steps shown in . always The desire to keep the subject centered within the camera frame takes precedence over other aesthetic goals. Unlike currently existing tracking systems, the present invention adjusts the camera position. Where in the central part of the camera frame, called the "fixed zone," do you move the subject? Also move it.

The stationary band controller has various system functions stored in the digital processor 46. It is realized by a number of variables representing the parameters of the system. of these parameters The value changes according to a predetermined tracking style defined by the user. . Processor 46 processes these variables to accomplish predetermined controller functions. .

For fixed bands, the amount of angular error allowed in each direction before the camera position is readjusted. A pair of angle values (LEAD LEFT and DEAD RIGHT) corresponding to a range ) is described by Initially, the immobile zone is symmetric about the center of the subject and (− THETA and +THETA), where the variable THETA is the angle value is a parameter of the tracking situation. The anterior angle range of the immobile band is THET Equal to twice the A value.

Angular misalignment along one axis between the tracked subject and the direction of the tracking device/camera Represents the difference value. The fixed band controller takes the error variable value and converts this variable into a fixed band. Remap to a new error variable corresponding to . For the purpose of the example, the error variable is on the horizontal axis represents the angular error along the HERR, which is also equivalent to the angular error along the vertical axis VERR. It can be expressed as follows.

In FIG. 6A, the dead zone (dcadzonc) has the value (DEAD LEFT.

DEAD RIGHT). Therefore, digital processor 4 6, if the ERROR value is still within the dead band, as indicated by decision step 96. Determine whether the If the value of the error variable is within the fixed band, proceed to processing step 94. The error variable is set to zero as shown below. The error variable value is within the fixed band. When the object is in motion, it is in a state where it comes out from either side of the immobile band.

Once it is determined that the error value is not within the fixed band as indicated by step 96 , the direction of the error is determined as indicated at decision step 98 . If the error value is to the left of the immobile band, then the DEADLEFT value is determined by process step 100. It is subtracted from the error value (Value or ERROR) so that Conversely , if the error value is to the right of the fixed band, D The EAD LEFT value is subtracted from the error value (ERROR value). In this way, 1 When the body exits from either side of the immobile zone, the angular error is caused by the body being in the camera frame. It is not how far from the center of It is converted by subtracting the edges of the immobile zone to describe the

Finally, the stationary zone is configured as a reduced response zone. In this configuration, The response for the system in a small response band, that is, the relationship between error and motor speed is , not zero gain (can be set to a very low gain. Such a configuration configuration is the case when the system is completely (rather than not moving) as in the case of a fixed band configuration. Gently re-center the camera relative to the tracked subject. But something like this The configuration requires very tight control over the motor 42.44 at low speeds. and a precision stepper motor rather than a typical analog motor controller. This can only be achieved through data. Furthermore, such a configuration The processing performed by the directional continuity controller and the motion controller is Make it complicated.

(Deadzone 5hirt Control er) The second process executed by digital processor 46 is shown in FIG. Implemented by the algorithm steps indicated by block 84 and shown in FIG. 6B. includes a fixed band shift controller. The subject leaves the fixed zone and the camera follows the subject. When it begins to track, the immobile zone shifts. The purpose of the fixed band shift is that the camera is THETA. The goal is to let the moving subject catch up. Many restrictions are initialized before they can be transferred. The variable MAXDEAD is THE Maximum angle at which one edge of the fixed band shifts toward the center of the camera frame in TA Represents the value obtained by adding a number. The MAXDEAD value is a function of motor speed. MAX The motor speed at which DEAD becomes greater than the minimum value THETA causes the dead band to deviate. This is the point where it is allowed. Motor speed where MAXDEAD is equal to (2・THETA) The band is biased enough for the tracking system to start leading the subject. This is the point that was moved.

The variable DEADSHI FT represents the maximum speed in degrees/second at which the dead band is deflected. and is likewise a function of motor speed. The value DEADSHI FT is typically Some small fraction of data speed is immobile. Before shifting the band, the variable values of MAXDEAD and DEADSHI FT are set to MO set to a function of the TOR5PEED value, which value is indicated in process step 104. (Represents the current value of the speed of the motor 44.

The shift of the fixed zone occurs as follows. As indicated by decision step 106, The E RROR value is compared to the right edge of the dead band DEΔD RIGHT. E.R. If the ROR value exceeds the right edge of the fixed band, then indicated by process step 108. The immobile zone is shifted to the left so that the This leftward shift is DEADSHI FT Subtract the value from both the DEAD LEFT value and the DEAD RIGHT value. This is done by If the ERROR value does not exceed the right edge of the fixed band, the decision step Whether the ERROR value exceeds the left edge of the dead band, as indicated by step 11O. A comparison is made to determine whether E RROR value exceeds the left edge of the fixed band If so, the fixed zone is shifted to the right, as indicated by process step 112. . This rightward shift of the fixed zone is the DEAD RIGHT value and the DEAD LEF This is done by adding the DEADSHI FT value to both T values.

The new values of DEAD LEFT and DEAD RIGHT are the shifted fixed bands. , which then indicates that the dead band has been shifted beyond the allowed limit of MAXDEAD. is compared with the MAXDEAD value to determine whether the To do this, judge As shown by step 114, the DEAD LEFT value is -MAXDE It is compared with the AD value. If the left edge of the immobile band exceeds the permissible limit, i.e. DE If AD LEFT<-MAXDEAD, the dead band is shifted to the left limit.

This sets DEADLEFT to -M, as indicated by processing step 116. Set AXDEAD and DEAD RIGHT to (2・T)IETA-MAX DEAD).

If the left edge of the dead band is within acceptable limits, as indicated by decision step 118. DEAD RIG) which is the right edge of the fixed band (T is the MAXDEAD value) compared to If the right edge of the fixed band exceeds the tolerance limit, the processing step As shown by 120, the immobile band has DEAD RIGHT to MAXDEA Set DEAD LEFT equal to (MAXDEAD-2・THETA ) is shifted to the right limit.

” To illustrate the process in point 1, the subject exits from the right side of the fixed band and the camera looks at the subject. Assume that you start tracking. Since the camera is in motion, the MAXDEAD value is Its minimum central value, before shifting the fixed zone in the opposite direction to the direction of camera motion. It is larger than the description THETA. If THETA is given a value of 5°, then the initial The immobile zone is (-5, +5)0. If the motor speed at that time is 25°/sec. , a typical value is 9° for the maximum allowed dead band value MAXDEAD, and the dead band deviation The maximum rate of DEAD SHIFT can be 1°/sec.

Over a period of 4 seconds, the immobile band shifts from (-5, +5) to (-9, +1).

At this time, the camera is not at the 5° that would have been the result if the fixed band had not been shifted. Therefore, it will lag behind the main body by 1°.

When the camera moves at a faster speed, the limit of the immobile band MAXDEAD will eventually become thicker than 10° (shifts, the fixed zone begins to shift to (<-10, <0), and is not tracked. It produces the effect of being in front of the subject. If the main body is slow (the maximum immobile band is MAX) DEAD is again reduced to 5° and the fixed zone is again at its normal position corresponding to its center position. begins to shift to rest values (-5, +5), making the subject symmetrical in the camera frame. Move it to a fixed area.

The size of the immobile band is 2・THETA), a function that controls the speed at which the immobile band is shifted. DEADSHT　FT and controls the amount by which the moving subject is preceded by the camera The function MAXDEAD all depends on the tracking method as defined by the user. It is a parameter that is adjusted. DEADSHI FT function selection is system control will affect the overall stability of your servo loop and will be obvious to those skilled in the art. Note that , must be chosen to avoid instability.

(Continuity Control) For human viewers, digital processor 46 to produce a more pleasing tracking style. A third process executed by is represented by processing block 86 in FIG. The directional continuity code achieved by the algorithm steps shown in Figure 60 controller included. Human viewers are very sensitive to changes in the direction of motion in images . Therefore, it is important to minimize the number of times the camera motion direction changes. ! C-wins occur at low speeds or when the tracking device ffi/camera is moving at high speeds. It is particularly sensitive to changes in direction that occur when moving from a stop to a stop and vice versa. direction Orientation Continuity The controller is used to create a series of directional changes that are unpleasant to the viewer. I guarantee that there will be no. Similar to the fixed band and fixed band deviation controllers, the direction coupling The continuity controller is stored in digital processor 46 and is used in various systems... It is implemented with a large number of variables representing parameters. These parameter values can be set by the user The tracking mode varies according to a predetermined tracking mode. processor 46 processes this variable to obtain the desired controller functionality.

The variables for the directional continuity controller are as follows. LAST DIR The ECT variable represents the sign of the direction in which the camera was last moving. ERRO RDIRECT is the sign of the angular error between the camera direction and the subject position, i.e. the camera represents the direction in which the object must move in order to approach the subject. 5TABL The E TIME variable represents the time in seconds that has elapsed since the last movement of the camera.

REVER8E DELAY variable allows the subject to change before the camera starts moving. Represents the minimum time required to move in a direction. REVER3E DEL The camera should change direction until the time specified by the AY variable is met. is not allowed. REVER5E Part of the DELAY time is mainly due to the camera frame. Note that it is spent when intersecting the immobile zone of the system.

The algorithmic steps performed by the directional continuity controller are shown in Figure 6C. is shown. LAST DIRECT and 5TABLE at any time The TIME variable specifies the direction of the last movement of the camera and the elapsed time since such movement. Each represents time. If another control process tries to move the camera in the same direction Then such movement is allowed and the elapsed time is reset. to do this As shown in decision step 122, processor 46 determines LAST DIR Compare the value of the ECT variable to zero. As shown in process step 126, LA The value of ST DTRECT is equal to zero indicating that the camera has not moved recently. , the camera is allowed to move in the wrong direction. In step 126, 5TAB The value of the LE TIME variable is set to zero and the value of the LAST DIRECT variable is set to the value of the ERRORDIRECT variable.

In decision step 122, the LAST DIRECT variable indicates that the camera has recently moved. If not equal to zero, indicating that the Then, the value of the ERRORDIRECT variable is compared to the LAST DIRECT variable. Ru. As mentioned earlier, the ERRORDIRECT variable If it is equal to the LAST DIRECT variable indicating that the direction of the camera is the same, Process step 126 is executed.

As shown in process step 128, the error direction and the last camera direction are If not equal, the elapsed time since the last exercise 5TABLE TIME is REVER 3E Compare with DELAY constant. 5TABLE TIME variable value is REVE R8E　If it is greater than or equal to the DELAY value, as mentioned earlier Then, process step 126 is performed. However, as shown in process step 130 , the elapsed time since the last exercise is specified in the REVER3E DELAY variable. If the time is less than the specified time, the camera is not allowed to reverse direction. processing stella , the value of 5TABLE TIME is incremented and the ERROR value is is set to zero. The value of the LAST DIRECT variable remains unchanged. Ru.

In this way, the directional continuity controller determines whether the camera moves in the direction of its recent movement. allow. However, there is a time delay before the camera is allowed to reverse direction. A minimum threshold of is required. REVER3E The time specified in the DELAY variable The delay value is determined by the tracking pattern as indicated by the user-defined 5TYLE signal. This is one parameter of the situation. REVER5E Typical range of DELAY variable values can cause instability in the servo loop of the tracking system. It can be from 1/6 of a second to longer than 5 seconds.

(Movement controller Cotton Controller) Digital Pro A fourth process executed by processor 46 is performed by processing block 88 of FIG. The motion control achieved by the algorithm steps shown in FIG. 6D is shown. Including ra.

In most video images, the background typically represents a larger portion of the image than the subject . If the movement of the scene relative to the background is rapid, even if this is due to the inconsistent movement of the subject. Even those resulting from traces can cause discomfort to the viewer. exercise control The roller has an acceleration time constant that is different from the deceleration time constant for the motor that drives the camera. The selection of numbers performs a pseudo-filter function on the system's response.

As a result, while avoiding the tendency to pore over the subject when the camera is stopped, , the camera movement starts smoothly. In addition to the filter function, the motion controller , limited to the maximum speed and acceleration of the camera depending on the situation of the subject's motion controller impose. As in other processing controllers, the motion controller uses digital A number of variables are stored in the file processor 46 and represent various system parameters. It can be realized. These parameter values correspond to the desired tracker selected by the user. Changes according to King Mode.

The variables associated with the motion controller are as follows. The variable ERROR is immutable Defined to be similar to variables used by the band controller. variable LAST ERROR represents the previous ERROR value. The variable MAXACC is the selected truck. represents the maximum permissible acceleration measured in degrees per square second for the locking mode.

The variable MΔXVEL is in degrees/square second for the selected tracking mode. Represents the maximum permissible speed measured. The variable K ACC indicates that the motor 44 is accelerated. An exponentially smooth time constant that controls the rate ngtime constant). The value (1-K ACC) is Exponential average history (cxponc) of the signal remaining after one time step ntia Ily averagezed bisLory).

The variable K DEACC is an exponential smoothness time constant that controls the speed at which the motor 44 is decelerated. represents a number. The value (1-KDEACC) is the value of the signal remaining after the one-time step. It is a fraction of the exponential average history. Typically, the velocity K DEACC is determined by the movement of the camera. Make acceleration as smooth as possible while avoiding overshoot mainly when stopping. The variable K is greater than or equal to ACC. K ACC and K D Both values of EACC are in the range between zero and one.

The motion controller operates as follows. The ERROR value is in tracking mode. If the camera speed is greater than the maximum allowable speed predetermined by the Speed is limited. This results in ERROR, as shown in decision step 132. This is done by comparing the value to the MAXVEL value. As shown in procedure step 134 If the ERROR value is greater than the MAXVEL value, the speed will be ERROR Limited by setting the value equal to the MAXVEL value. Step 132 If the ERROR value is not greater than MAXVEL, decision step 136 This is compared to the -MAXVEL value, as shown in . ERROR value is -M If it is less than the AXVEL value, the velocity is re-set as indicated in process step 138. and ERROR values equal to -MAXVEL.

LAST ER is the difference between the current value of E RROR and the previous value of E RROR ROR (acceleration) is the maximum acceleration determined in advance by the tracking situation. If the value is larger, the camera speed is the camera speed at that time plus the maximum acceleration. limited. To this end, as shown in decision step 140, the ERROR The value is compared to the sum of LAST ERROR and MAXACC. ERROR value is large in step 140, as shown in process step 142, By setting the ERROR value to the sum of LAST ERROR and MAXACC. limited. Otherwise, the ERROR value is indicated in decision step 144. It is compared to the current velocity minus the maximum allowable acceleration. E The RROR value is L A S T E RROR minus MAXACC. If so, the acceleration reduces the ERROR value to L, as shown in process step 146. By setting ASTE RROR minus MAXACC and limited.

After limiting the acceleration or velocity, the digital Le processor 46 determines whether the camera is accelerating. camera added If the vehicle is running at high speed, smoother acceleration will be obtained, as shown in process step 150. . Smooth acceleration is determined by the ERROR value and the final smoothed error by equation (1) below. By calculating the smooth error value N5ERROR using the value of LSERROR. will be achieved.

NSE■0R=(+-K ACC)・1ERROR+to ACC・ER1? ORC Equation 1) Processing step 150 is equivalent to low-pass filtering. Next, equation (1) The new smoothed error N5ERROR is used to control the speed of the motor 44.

If the camera is not accelerating, the processor 46 determines whether the camera is decelerating. If the camera is slowing down, As shown in process step 154, a smooth deceleration occurs. Smooth deceleration is E　Error LSERR(' ) is achieved by calculating the smoothed error value using the value of H.

N5El? ROM-(DEACC to +-)・I-(iERROR+K rlEA cc+ERROR (Equation 2) The above pseudo low-pass filtering method is used during acceleration and deceleration. Use different time constants during 4 Viewers' sensitivity to the smoothness of acceleration is 1, in a preferred embodiment, K ACC The ratio between KDEACC and K DEACC is typically 12.

In a first embodiment of the invention, the user of the tracking system Specify the tracking situation so that performance is specifically tuned to the dominant movement situation do. Typical selections of modes or tracking styles are Slow, General and and high speed operating motes. Lectures, slow sports, children's games, or The low-speed operation mode, which is suitable for movements that do not involve running, is when the subject is in a state of rapid movement. If so, it is characterized by slow, smooth motion that tends to lag behind the main body. general movement A general mode of operation suitable for Suitable for smooth movements with a tendency to move. Tennis, basketball/pole, soccer Alternatively, the high-speed operation mode is suitable for running at full speed and movements with sudden changes in movement. , even the fastest moving objects are suitable for tracking; Tendency to respond more quickly than is visually pleasing if not fast enough to cover this has.

(Automatic Context Dete According to the second embodiment of the present invention, A user-supplied 5TYLE signal describing a given tracking style is as described by recent values for some of the system variables, as described below. Based on the past history of the motion of the subject being tracked, the digital processor 46 is replaced by a signal generated by

FIG. 7 shows a digital processor 4 for processing signals VERR and HERR. 6 shows the process executed by 6. In this example, the tracking system The response characteristics of the video give the resulting attributes that are subordinate to the human viewer. Processor 46 automatically adjusts the process described above so that the process is performed. In Figure 7 , the access point indicated by block 78.80.82.84.86.88.90 The algorithmic process is similar to that described with respect to FIG. 5 and the first embodiment of the invention. are doing.

The fifth process executed by digital processor 46 is shown in FIG. Implemented by the algorithm steps represented by lock 92 and shown in FIG. Contains automatic situation controllers.

Part of the functionality of a tracking system is to determine the location of the tracked subject. This provides a wealth of information that can be used to improve your future exercise style. You can get information. Typically, the actual movements performed by the subject are very small within their range of motion. stable. In a second embodiment of the invention, the parameters of the subject's motion are stored. and then compared with parameters expected to be related to tracking conditions etc. compared. The extracted parameters are based on the special characteristics of the movement and the appropriate situation at that time. Includes the percentage of time to track, the maximum angular acceleration of the subject, and the minimum angular velocity of the subject.

These parameters were chosen for ease of calculation rather than as best parameters. be done.

Automatic context determination controller The variables related to the minination controller are as follows. . The CON T E X T variable is equivalent to the 5TYLE signal, and the indicates the status or mode of tracking. The MISSTRACKING variable is It represents the length of time that the subject is outside the immobile zone. The MAXMISSTRACK variable is , for a particular tracking style selected by the CONTEXT variable. Represents the maximum permissible percentage of time that a subject is allowed to be outside the immobile zone. A.V. The GMOVE variable represents the average time the subject moved in one direction. MOTO The R5PEED variable is the same as defined previously for the first example.

The DELTA ERROR variable indicates the degree unit since the subject's angular error was last measured. represents the change in the angular error of the position. DELTA ERROR is MOTOR5P EED is added to yield the subject's velocity CURRENT 5PEED. DEL The TA5PEED variable is the unit of 0/second since the subject's angular error was last measured. CURRENT represents the change in 5PEED. MAX 5PEED variable does not represent the maximum angular velocity of the subject. The MAX ACC variable is the maximum angular acceleration of the subject. represent.

Following processing by motion controller 88, automatic situation determination controller 92 It is determined whether the tracking situation at the time is appropriate for the movement of the subject.

In Figure 8, each new part of ERROR changes the current system mode. The number of hours the subject was unsuccessfully tracked is calculated as indicated in process step 160. be done. E. Use the absolute value of the RROR variable to determine whether the subject is currently in a tracking failure state. Determine whether or not. F, Is the absolute value of the RROR value smaller than the maximum allowable error? is equal to this, the MI5STRACKING value is not incremented and the entity Indicates that the patient is still moving within an acceptable range. However, ERROR If the absolute value of the value is greater than the maximum allowable angular error, then the MISSTRACKING value is Indicates that data is incremented from a range acceptable to the subject.

The percentage of time system 15 is in a failed tracking state is indicated at decision block 162. Calculated by dividing the MT5STRACKING value by the total time, so that and the maximum allowable percentage and ratio to the mode MAXMI 5STRACK at that time. compared. If the system is in a tracking failure state, as shown in process step 164, If the percentage of time is 5STRACK, then the MISSTRACK value is MAXMT 5STRACK The situation where the value of the resulting C0NTEXT variable is the next fastest, i.e. tracking mode. The maximum angular velocity of the subject is then indicated in process step 166. calculated as follows. This is done by calculating the MAX 5PEED value using the following formula (3) etc. The final value of MAX 5PEED and the subject's current speed CU, calculated at RRENT　By setting it to the value with the largest absolute value of 5PEED. be exposed.

CIIRRENT　Sr’EED=(MOTORSl’EED+DELTA　E RROR) Formula (3) The absolute value of the speed of the main body is MAX 5 PEED value depends on the direction. used as it exists.

Next, the maximum angular acceleration MAX ACC of the subject is indicated by processing block 168. (Calculated. This is the MAX ACC calculated in the following formula (4) etc. The final value and the absolute value of the subject's acceleration at that time CURRENT The largest value of ACC This is done by setting the MAX ACC value.

CURRENT ACC=DELTA Sl’EED (Formula 4) Main speed difference The absolute value is used so that the value of MAXACC is direction dependent.

Finally, the average value of the subject's motion is calculated as shown in process step 170. Ru. The average time is calculated by calculating the average time between changes in direction and adding the AVG MO to this average value. It is calculated by simply setting the VE variable.

Each resulting situation, or mode, is defined by the range of appropriate motion of the subject and the corresponding range of values. It's connected. As indicated at decision block 172, A determination is made whether the value at time more closely matches the value for the slower situation.

If the characteristics of the subject's motion closely match the relatively slow mode or situation, the processing step As shown in step 174, the CONTEXT value is set to the next slow situation. will be played.

Otherwise, the tracking status remains the same.

(Variable Ranges) Tracking system implementation Various other variables may be relevant, depending on the actual physical configuration. For example, the GAIN variable is required to control the overall gain of the servo loop in a tracking system. The key point. The MOTOR0FFSET value is set by the system when using analog motors. Allows the system to take into account the required minimum starting voltage. motor control This is done by adding the MOTOR0FFSET value to the ERROR value during the process. The zero-crossing response of a motor such as can be balanced.

Setting the MOTOR0FFSET variable to zero is a This produces an effect that adds to the immobile band in a way that is difficult to compensate for with variables.

The entire tracking system disclosed in any embodiment of the invention can be integrated into a single system. are included in the parameter set for each different mode to form a turbo system. It is difficult to give actual values for the variables. Such parameter values are , highly dependent on the specific considerations of the entire electromechanical system. However, the following between slow-acting tracking mode and fast-acting tracking mode. There are many ratios that present reasonable ranges for the selected variables. tracking system The overall gain of the system, i.e. the response of MOTOR5PEED to ERROR, is 4 over a constant range of 1. The ratio of K ACC to K DEACC is constant at 1:3. over a range of numbers. The values of K ACC and K DEACC are respectively 3:l constant. over a range of numbers. REVER3E DELAY time ranges from 1/6 to 6 seconds . Finally, the value of THETA spans a constant range of 5·1.

(Another embodiment) The present invention is described in an exemplary ultrasonic tracking system using a DC motor. However, those skilled in the art will understand how the present invention applies to any tracking system. It will be clear that it is equally applicable.

Such a system may be active, as in the embodiment, or passive tracking systems such as infrared tracking systems that do not require a transmitter. It can also be a locking system t1. Other such tracking systems are , camera frames using frame IIi motion comparison and pattern recognition techniques. includes a system for identifying portions of the image to be maintained within the system. Still other systems may be sampled or continuous, and infrared, Communication between transmitter/receiver can also be done using radio frequency or sound wave transmission. . Other systems use cameras or motors instead of analog motor controllers. A precision stepper motor may be used to control the tracking device.

” In the embodiment of item 1, blocks 78.80.82.84.86.88.9 The algorithm processing described by 0.92 takes into account multiple system variables and It is realized by a programmed digital processor. For those skilled in the art, small, medium By using conventional or large logic components as well as analog circuitry, It will be clear that a functionally equivalent process can be achieved. hard like this ware configurations are well within the purview of those skilled in the art and within the scope and spirit of the invention. It is in.

For the present invention, the horizontal angular error signal HER representing one axis of a multi-axis system is Described in relation to R. Those skilled in the art will appreciate that the designer understands the human response to vertical axis motion. Taking into account the different sensitivities of the eyes, the control function disclosed in the main text will reduce the vertical error signal. It will be clear that it is equally applicable to No. VERH. For those skilled in the art, Digital processor 46 detects horizontal and vertical angular errors as shown in FIG. be able to perform control functions for both signals, or for two such signals. It is further clear that digital processors can be used, one for each axis of motion. Dew.

Furthermore, since many motions have different motions in each axis, our tracking system The stem can be assigned different tracking modes for each axis. Ru. Conversely, the present tracking system supports horizontal and vertical axis tracking. The parameters of these modes produce a specific two-axis motion. automatic or user-selected tracking modes defined to can be provided.

Furthermore, the inventive concept can be applied to a tracking system given a set of criteria. Optimize system performance to meet this criterion using specific parameters of It should be understood that it applies to any tracking system. Example listed a number of visual criteria. Those skilled in the art will appreciate that given the appropriate system parameters. If possible, other information not disclosed in the main text but related to the subject's smart tracking. It is clear that the criteria of are similarly achieved by the apparatus and method disclosed in the present invention. Or maybe. Optimize the tracking system performance to meet the above criteria. Such criteria and means for determining are within the scope of this invention.

Therefore, it is clear that the detailed disclosure has been made by way of example only and not with the intention of limitation. Will. Various changes, modifications, and improvements will readily occur to those skilled in the art, and the present invention It may be practiced without departing from the spirit and scope of the Act. The present invention It is intended to be limited only by the claims and their equivalents.

Shi- FIG. 5 FIG, 6C Submission of translation of written amendment (Article 184-8 of the Patent Act) 1993 3Jl 5th [Tsukasa

Claims (22)

[Claims] 1. A system that allows a device to track the movement of a moving subject Leave it behind. a. a signal source that generates a signal that can be tracked; b. Trusted from the above source Continuous analog function with changes in bearing relative to the signal control means for generating associated orientation data in the form of a control signal that varies as; c. selected values each processing said control signal according to the expected movement of the subject; means for storing a plurality of sets of data values representing; d. Select one of the data values in the set c. according to one of said set of data values selected by said selection means. means for processing said control signal to produce a processed control signal; f. Said processing moving said device about at least one axis in response to a control signal generated by said device; A system equipped with a drive device that keeps the device facing the subject. 2. each of said data value sets for a specified range of expected movements by the subject; , a data value representing a selected acceleration of said device and a selected deceleration of said device. 2. The system of claim 1, comprising: 3. the set of data values corresponds to a particular range of expected movement by the subject; the selected angular velocity of said device and the maximum permissible angle between the axis and main body of said device; 2. The system of claim 1, further comprising a data value representing an error. 4. The selection means indicates which of the plurality of sets of data values is to be selected. 4. A system as claimed in claim 1, 2 or 3, including user defined signals. 5. The selection means selects a future relative movement of the subject based on the past movement of the subject. 2. The method of claim 1, further comprising means for automatically selecting one of said plurality of sets of data values in anticipation. , 2 or 3. 6. A method for intelligent automatic tracking of moving subjects with cameras There, a. detecting the last encirclement that the camera moved around the axis; b. camera finally moves calculate an elapsed time value that represents the time interval that has elapsed since c. determining an error direction representative of the subject's current orientation with respect to the camera; d. Before comparing the recorded error direction with the last direction; e. the last direction and the error direction If they are equal, move the camera around the axis in the error direction, f. If the last direction and the error direction are not equal, predict the elapsed time value. Compare with the determined value, g. If the elapsed time value is at least equal to a predetermined value, then the camera is – direction around the axis A method that includes steps. 7. h. If the elapsed time value is less than the predetermined value, the camera 7. The method of claim 6, further comprising the step of maintaining an azimuthal angle. 8. The predetermined value corresponds to the period of direction change by the subject for the predetermined type of movement. 7. The method of claim 6, wherein the method is selected to represent an average time of . 9. A device that intelligently and automatically tracks a moving subject using a camera. Leave it behind. a. a method for determining an error direction representing the current orientation of the subject with respect to the camera; step by step, b. means for detecting the last direction in which said camera moved about an axis; c. camera is means for calculating an elapsed time value representing an elapsed time interval since the last exercise; d. If the error direction and the last direction are not equal, the last direction is the error direction. means for comparing and comparing the elapsed time value with a predetermined value; e. The elapsed time value is small. at least equal to a predetermined value, or said last direction is equal to the error direction. If so, the camera is moved forward in the error direction in response to the determining device and the comparing means. A drive device for causing movement around one axis. 10. The determining means receives a source signal from the subject and determines at least one aspect of the source signal. From the signal along 10. The apparatus according to claim 9, further comprising control means for generating. 11. Intelligent and automatic tracking of moving subjects within the camera frame In the method of doing a. moving said camera axially to place a subject within a camera frame; b. establishing predetermined limits for angular velocity and angular acceleration about the axis of the camera; , c. a central region within the camera frame that is substantially symmetrical about the center of said subject; define, d. moving the camera around an axis at a speed not exceeding the predetermined limit; holding the body within the area; e. represents the total time that the subject was outside the central region in a given time interval. Calculate the first percentage, f. calculating the maximum angular velocity and acceleration of the subject within said given time interval; g. Before the prediction for the angular acceleration and angular velocity of the camera according to the expected movement of the recording subject; Redefine established limits A method that includes steps. 12. step (g) comparing said first percentage with a predetermined maximum allowable percentage; and when the first percentage exceeds an established predetermined limit, the angle of the camera is including selecting new predetermined limits with incremental values for velocity and angular acceleration. 12. The method according to claim 11. 13. step (g), wherein the maximum allowable angular velocity and maximum angular acceleration of the subject are the When the angular velocity and angular acceleration of the camera are smaller than the predetermined limits, the angle of the camera including selecting new predetermined limits with decrement values for velocity and angular acceleration. 12. The method according to claim 11. 14. Place a moving subject within the camera frame in a way that is visually pleasing to the viewer. In an apparatus for permitting a camera to hold: a. the camera around the axis means for storing a plurality of predetermined limits for the angular velocity and angular acceleration of; b. responsive to said storage means, a predetermined limit for retaining the subject within said area; a drive for moving said camera about an axis at a speed not exceeding c. the turtle a hand defining a central region substantially symmetrical about the center of said subject within a frame; step by step, d. represents the total time that the subject was outside the central region in a given time interval. a first means for calculating a first percentage; e. of the subject within the given time interval. a second means for calculating maximum acceleration and maximum angular acceleration; f. In response to the first and second means, the subject moves forward in anticipation of the subsequent movement of the subject. new predetermined limits from said storage means for the angular acceleration and angular velocity of said camera; a means of selecting A device equipped with 15. The selection means selects the first percentage from a predetermined percentage stored in the storage means. The first percentage exceeds a predetermined maximum allowable percentage. How to select new predetermined limits for camera angular velocity and angular acceleration when 15. The apparatus of claim 14, comprising a stage. 16. The selection means selects the maximum angular velocity and maximum angular acceleration of the subject from the angle of the camera. The maximum angular velocity of the subject compared with the then predetermined limits for velocity and angular acceleration. The maximum angular acceleration is substantially less than or greater than the predetermined limit for that samurai. New predetermined limits for the camera's angular acceleration and angular velocity when substantially large. 15. The apparatus of claim 14, including means for selecting. 17. Intelligent and automatic tracking of moving subjects within the camera frame In the method of doing a. substantially symmetrical about the center of said subject along one axis at a first time; defining a central region within the camera frame by a pair of angular coordinates that are the target; b. No. determining the angular error between the subject and the camera along the one axis at a time of 2; c. determining whether the angular error is within a central region; d. The angle error is medium If not within the heart region, the camera reduces the angular error at the second time. determine the error direction representing the direction in which it must move; e. moving the camera about the one axis in the error direction; f. Before The angular coordinates of the central region are shifted in the direction opposite to the error direction. redefine A method that includes steps. 18. Step (f) is f. 1 Subtract one value of the angular coordinate of the central region from the angular error value and calculate the difference value. form, f. 2 Subtract the difference value from each of the angular coordinates to form new angular coordinates. 18. The method of claim 17, including the step of: 19. g. If the angular error is within the center area, then the camera is 18. The method of claim 17, further comprising the step of maintaining the orientation. 20. Intelligent and automatic tracking of moving subjects within the camera frame In the device that performs a. substantially symmetrical about the center of said subject along one axis at a first time; means for defining a central region within the camera frame by a pair of angular symbols; , b. creating an angular error between the subject and the camera along the one axis at a second time; a means of defining; c. means for determining whether said angular error is within a central region; d. the angle If the error is not within the center region, the camera detects the angular error at the second time. means for determining an error direction representing the direction in which the movement should be made in a decreasing manner; e. means for moving the camera about the one axis in the error direction; , f. the angle of the central region such that the central region is shifted in a direction opposite to the error direction; a means of redefining the coordinates; A device equipped with 21. The redefining means The difference is obtained by subtracting one value of angular coordinates defining the central region from the value of the angular error. means for forming the value of; means for subtracting a difference value from each of said angular coordinates to form a new angular coordinate. 21. The apparatus according to claim 20. 22. Introduce a moving subject within the camera frame in a way that is visually pleasing to the viewer. In an apparatus for permitting a camera to hold: a. signals that can be tracked a subject that generates; b. receiving a signal from the subject, and generating a control signal from the signal; control means for continuously generating relative azimuth data of the representing the angular error between the positions of said subject and camera along at least one plane; c. Before means for anticipating future relative motion of the subject based on past motion of the subject; d. Before When the expected movement is imparted by said means in response to said control signal, said control Determine whether the angular error represented by the signal is within an acceptable range and proceed. means for processing the control signal if the angle error is not within an acceptable range; e. selectively moving the camera along an axis in response to the processed control signal; and a drive device for holding the subject within a camera frame. .