JPH07120241B2 - Vector input device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Outline] In a device capable of inputting various kinds of vectors by detecting the direction and amount of pressure applied to an input plate by a pressure sensor, a disturbance such as vibration is applied to the input plate. When added, this is detected by a combination of a pseudo input plate and a pressure sensor to obtain a true vector input.

[Industrial application field]

The present invention relates to a vector input device capable of inputting a vector amount by minutely moving one operation input plate in multiple directions.

[Conventional technology]

For electronic devices that require character input, a device that can input the character pattern itself may be used instead of the keyboard because of the ease of input. For example, an example is a matrix-shaped character input plate in which switches (sensing points) that are turned on and off with minute strokes are arranged in each grid-shaped block. When the surface of the character input plate is traced with a fingertip, the switches of the pressed portion are sequentially turned on, so that the input pattern can be discriminated from the order.

[Problems to be solved by the invention]

However, in the above-mentioned sensing point method, in order to improve the resolution, the number of switches increases, the structure becomes complicated, and the cost becomes high. There is a method of forming a grid-like sensing point by using a plurality of pairs of infrared ray emitting / receiving pairs instead of switches, but the apparatus becomes large and expensive.

The present invention uses one operation input plate that can move in a plane and a pressure sensor that can detect the direction and amount of the force applied thereto as an input unit, so that various types of vector quantities required for pattern configuration can be obtained. When a disturbance such as a vibration is applied to the input plate, the disturbance can be separately detected and a true vector input can be obtained.

[Means for solving problems]

FIG. 1 is a basic configuration diagram of the present invention, in which 1 is a pressure sensor, 2 is an amplifier (for example, operational amplifier) used as necessary,
3 converts its output (analog value) to digital value
A / D converter, 4 is a microprocessor (MPU) or digital processing processor (DSP) that detects an input pattern from the input digital value and compares it with a reference pattern to perform determination processing, and 5 is a pattern storage memory Is.

There are two types of pressure sensor 1. The first is sensors S _{1 to} S ₄ that detect the direction and amount of the force applied to the operation input plate for vector input, and the second is the same as the operation input plate for detecting disturbance such as vibration. Sensors S ₁ ′ to S ₄ ′ having a mass and detecting the direction and amount of a force applied to a pseudo input plate which is provided so as to be capable of minute movement in a plane parallel to the input plate.

[Action]

An arbitrary vector can be input by combining the operation input plate and the pressure sensors S _{1 to} S ₄ . However, when vibration is applied to the input plate, the sensors S _{1 to} S ₄ produce outputs although no operation is performed.
Processing this as a regular vector input causes a malfunction. Therefore, in the present invention, the sensor S ₁ ′ on the pseudo input plate side is
The true vector input is obtained by removing the vibration component by taking the difference with the output simultaneously generated in ~ S ₄ ′.

Such a vector input device can input a lot of information with one input plate. One is the direction of force, and the amount of force is also powerful information. Furthermore, the temporal change of the vector is essential for inputting the character pattern. To give a simple example, when pressure is applied to the operation input plate according to the pattern of the character to be input, one character is input if the amount of force does not matter, but for example, it is strong even if pressure is applied in the same direction. It can be determined as another character pattern depending on whether it is weak. For example, when the number "3" is input, if the pressure is strong, it is determined to be uppercase, and if it is weak, it is determined to be lowercase. Of course, it is not limited to such clear character patterns. For example, it is possible to selectively use the process A when the pressure is strong and the process B when the pressure is weak even if the pressure is applied in the same direction.

〔Example〕

FIG. 2 and subsequent drawings are explanatory views of an embodiment of the present invention. FIG. 2 is a structural diagram applied to an on-vehicle audio device. 6 is a device housing, 7 is a common display, 8 is an individual display, 9 is an operation input plate, 10 is a spacer, and S _{1 to} S ₄ 4 input plate sensors, 11
Is the lid.

The input plate 9 includes four-way eyeways that individually apply pressure to the sensors S _{1 to} S ₄ , two adjacent sensor pairs (S ₁ , S ₂ ),
(S ₂ , S ₃ ), ... can be moved minutely in four directions by applying pressure simultaneously. The spacer 10 is for preventing the load of the operation plate 9 from being applied to the lower sensor S ₁ when it is not operated. Inside the housing 6, a pseudo input plate 12 having the same mass as the operation input plate 9 and capable of minute movement in a plane parallel to the input plate 9 is provided. The sensors S ₁ ′ to S ₄ ′ are for detecting the direction and amount of the force applied to the pseudo input plate 12, and the sensor S _i ′ (i = 1 to 4) is the amount of force in the same direction as the sensor Si. To detect. When the pseudo input plate 12 is actually formed, the window 13 may be formed in order to facilitate the insertion of the substrate or the like inside the housing. At this time, in order to make the mass equal to that of the input plate 9, for example, if the material of the input plate 9 is an acrylic plate,
The material may be selected such that the pseudo input plate 12 is made of aluminum die cast. Further, by doing so, the latter can be made thinner and the space inside the housing can be effectively utilized.

The outputs of the respective sensors S _{1 to} S ₄ and S ₁ ′ to S ₄ ′ are amplified by the amplifying section 2 as shown in FIG. 3 and then input to the A / D converter 3. This A / D converter 3 selects only one of the 8 input channels ch1 to ch8 by the 3-bit selection signal SEL and A / D
D-convert. SO is an A / D converted digital value (for example, 1
It is a serial output (sample 8 bits), and CLK IN is a clock input for transfer. Control of A / D converter 3 is MPU4
Done by This MPU4 is a 3-bit selection signal SEL
Output OUT0 to OUT2, transfer clock output CLK OU
T, A / D conversion data serial input SI, other control output OU
Equipped with T3 to OUTn.

FIG. 4 is a specific example of the pressure sensor unit 1 and the amplification unit 2. The pressure sensor unit 1 is an equivalent bridge-connected variable resistor VR.
_A constant current is made to flow from the reference section 21 to the pressure sensitive element represented by _{1 to} VR ₄ , and the output currents flowing to the terminals 22 and 23 are inputted to the amplification section 2. The amplification unit 2 includes an operational amplifier OP ₁ for amplification having a 1 / V conversion function and an operational amplifier OP ₂ for level shift, and generates a linear output voltage with respect to pressure (however, only in the positive direction). . However, since the characteristics of the sensor pair Si and Si ′ do not necessarily match, the difference and the offset amount are corrected in the difference calculation as described later.

The sampling by the A / D converter 3 is continuously performed a plurality of times for the same channel. FIG. 5 shows an example in which sample Nos. 0 to 40 are sampled 41 times continuously at the sampling interval T. By doing so, it is possible to detect a pattern of temporal changes in the amount of pressure received by the same pressure sensor. A / D conversion value is 8 bits, taking a value from 00 _H to FF _H. The A / D conversion value for this sample No. 0 to 40 is 1
One input pattern is stored below the address #xxxx in the memory map of FIG.

A basic pattern to be compared with the input pattern is registered in the addresses # 0000 to # ΔΔΔΔ of this memory up. For example, the pattern of M ₁ is registered in the addresses # 0000 to # 00BF. This is the operation input plate 9 in FIG. 2 (b).
This is the standard value of the pattern obtained when is pressed strongly in the A direction. In this pattern, only the sensor S ₁ produces an output and the remaining sensors S _{2 to} S ₃ do not produce an output, so the basic pattern of M ₁ is as shown in FIG. 7 (a). The A / D converted value is stored in the memory map of FIG. 6, so the sensor S
Valid values 3E, 6C, 99, ... Appear in the category corresponding to ₁ , but
The A / D conversion values of the sections corresponding to the sensors S _{2 to} S ₄ are all 00 (indicated by − in the figure). Regarding the basic pattern of M ₂ , effective pressure 3E, 6C, 99, ... appears only in the output of sensor S ₂ because pressure is applied in the direction of Fig. ₂ , but the remaining sensors S ₁ , S
The output of ₃ and S ₄ is 00. FIGS. 7 (a) to 7 (d) show sampling patterns of the basic patterns M _{1 to} M ₄ obtained in this way, and the arrows attached to the lower side of M _{1 to} M ₄ are shown in FIG. b) Each direction eyeway is shown.

When the operation input plate 9 is moved in the diagonal direction in FIG. 2B, two adjacent sensors produce outputs simultaneously. This standard value is the combination pattern of FIG. 9th
FIG. 10 and FIG. 10 show standard values of patterns generated when the operation input plate 9 is operated weakly. For example, the basic pattern m _{1 in} FIG. 9 (a) is an operation in the same direction as in FIG. 7 (a), but the pressure is weak, so the A / D conversion value is lower than that in the basic pattern M ₁ . Similarly, the combination pattern m _{12 in} FIG.
Although the operation is performed in the same direction as in FIG. 6A, the pressure is weak and the A / D conversion value is lower than that of the basic pattern M ₁₂ .

The examples of FIGS. 7 to 10 described above are basic vector patterns, but a normal character pattern is a combination of these. FIG. 11 is an explanatory view of this, and (a) to (c)
Is each pattern of numbers "1" to "3". While these are one-time learning patterns, (d) is a five-time learning pattern. When the number of times of learning (for example, the average) is increased in this way, the quality as the reference pattern is improved. It should be noted that what is to be learned may be indicated by a separate switch operation. These reference character patterns are also registered in the memory map shown in FIG. 6, and when the input pattern is actually taken in, it is collated and the judgment result is obtained.

This process is shown in FIG. The same routine (A) is the main routine "MAIN", and the A / D conversion processing routine "ADT" and the timer interrupt processing routine "TINT" are shown in the same routine (B). Further, FIGS. 7C and 7D show each subroutine of the pattern determination processing "PAT" and the pattern matching processing "PATM".

The main routine "MAIN" starts from "Initialize". This processing includes timer interrupt cycle (sampling cycle) set, each flag reset, and the like. When this initialization processing is completed, (B) A / D conversion processing "AD
Move to T ". This selects the A / D conversion start ch and performs only one sample. At first, it is the sample No. 0 of channel ch1, and its AD value (A / D conversion value) is put in the register ADV. hereinafter similarly obtaining the AD value of the sample No.0 channel Ch2～8. However, the channel ch5~ch8 (I1 = 5~8) is because they are related to the output of the sensor S ₁ '~S _4', the sensor S ₁ ~ S ₄
And the AD value (I1 = 1 to 4) of. At this time, in order to match the characteristics of the sensor pair Si, Si ′, the AD values of the sensors S ₁ ′ to S ₄ ′ are corrected using two types of coefficients kc, ko. kc is a sensitivity adjustment coefficient, which corrects the inclination of the characteristic shown in FIG. ko is an offset adjustment coefficient, which adjusts the offset amount of the characteristic shown in the figure. The difference calculation formula using this correction coefficient is ADV (I1-4) ← ADV (I1-4)-(ADV (I1) * kc (I1
-4) + ko (I1-4)). For example, when I1 = 5, ADV (1) ← ADV (1)-(ADV (5) * kc (1) + ko
(1)) Then, the timer interrupt flag TF is set to 0 and the process returns to the main routine of (A). This flag TF is (B)
It is set by the timer interrupt routine "TINT" of 1 and when this is 1, A / D conversion processing "ADT" is performed.

In the processing of the main routine of (A) following this processing, the contents of the register ADV (*) are put into the old register ADO (*). * Indicates 1 to 4 corresponding to the sensor No. (channel No.)
ADO (*) ← ADV (*) is ADO (1) ← ADVJ (1), AD
Represents O (2) ← ADV (2), .... Then again (B)
Execute the A / D conversion processing routine "ADT". This is to detect a change in sensor input. In other words, this AD value
ADV (*) is put in the new register ADN (*), and the difference ADN between the two
If (*)-ADO (*) exceeds a certain value kAD, there is a change. If there is no change, replace ADN (*) with ADO (*) and return to, but if there is a change, index register
Put #XXXXX in IXO. This is the start address for storing the input pattern in FIG. From this start address
There is an address for storing the AD value ADO of each sensor (1) ~ADO (4) every 30 _H. Next, when the value of the index register is set to +1 (IXO ← IXO + 1), the storage address of sample No. 1 is taken as the point. After this, the new register ADN
Put the contents of (*) in the old register ADO (*) and store the input pattern up to sample No. 40 in the memory map.

In the next pattern determination process "PAT", as shown in detail in FIG. 12 (C), a standard pattern is extracted from the start address 0000 _H of the recognition pattern in the memory map, and it is subjected to pattern matching in FIG. 12 (D). Match with the input pattern in the process "PATM". If they match, the pattern matching flag PS
Since 1 becomes 1, the pattern No. at that time is put in the variable P. If they do not match, PS = 0 remains, so the standard pattern start address MPC is increased by CO _H. This means the reading of the next standard pattern. If there is no corresponding pattern even after repeating this to the final address # ΔΔΔΔ of the standard pattern, the process returns to the main routine with P = 0. In the main routine, each process is performed according to the value of P. Taking a cassette deck as an example, process 1 is PLAY, process 2 is FF, and process 3 is REW. The main routine of this example is an example in which the time required for A / D conversion is very short compared to the sampling period (T).

The pattern matching process "PATM" is performed using constant allowable values mk and Ek as shown in FIG. mk is a permissible value of the deviation from the standard pattern, and this is the upper and lower widths allowed when the input pattern as shown in FIG. 5 is determined to match a certain standard pattern. On the other hand, the allowable value Ek of the pattern mismatch error is the number of errors (value of the counter EC) which is determined as a match if a part of the input pattern exceeds the allowable value mk and is below a certain number. As described above, the readability can be improved by using the double allowable value.

FIG. 13 shows a specific example of removing the vibration component. FIG. 7A shows the output changes of the sensors S _{1 to} S ₄ when the numeral “3” is input. At this time, vibration as a disturbance is also added, and its component appears in the output of each sensor S ₁ ′ to S ₄ ′ shown in FIG. The figure (c) is a coefficient of the AD value of (b).
It is true input data obtained by correcting with kc and ko and subtracting the value from the AD value of (a). The readability is improved by performing the pattern matching based on the true input data.

〔The invention's effect〕

As described above, according to the present invention, since the direction and amount of the force applied to the operation input plate are used as input information, multiple inputs can be made. Therefore, the space in front of the device can be effectively used and the degree of freedom in design is improved. Especially for in-vehicle use, it is not necessary to check the switches, so the safety is improved. Further, by adding time and movement order in which force is further applied to the above direction and amount, it becomes possible to input a character pattern or the like.

Furthermore, even if external disturbance such as vibration is applied to the operation input plate, it is detected by the pseudo input plate and removed by signal processing. However, there is an advantage that it can be input without being affected by vibration.

[Brief description of drawings]

1 is a basic configuration diagram of the present invention, FIG. 2 is a structural diagram of an embodiment of the present invention, FIG. 3 is a block diagram of the same embodiment, FIG. 4 is a partial circuit diagram of FIG. 3, and FIG. Is an explanatory diagram of a sampling pattern by A / D conversion, FIG. 6 is an explanatory diagram of a pattern storage memory, FIGS. 7 to 11 are explanatory diagrams of various standard patterns, and FIG. 12 is a process performed by the MPU of the present invention. Flow chart, first
FIG. 3 is an explanatory diagram of a process when "3" is input.

Claims (2)

[Claims] 1. An input plate for operation, which can be finely moved in a plane,
A pressure sensor for detecting the direction and amount of a force applied to the moving direction of the input plate; a pseudo input plate having the same mass as the input plate and capable of minute movement in a plane parallel to the input plate; A pressure sensor that detects the direction and amount of force applied to the pseudo input plate, and A / that converts the outputs of these sensors into digital values.
A D converter, and a processor for determining the difference between the outputs of the sensors in the same direction of both input plates from the output of the converter and operating the operation input plate to determine the type of the input vector. A vector input device characterized by: 2. The vector input device according to claim 1, wherein the processor has a function of discriminating a character pattern from a temporal change of an input vector.