JPH04332866A - Noncontact driving distance/velocity measuring device - Google Patents

Abstract

PURPOSE:To eliminate a measurement error of a space filter for brightness due to solar light and lighting by compensating for change in luminance of an image accompanied by driving by using a deflection of an output of two grating-shaped space filters with different phases. CONSTITUTION:A signal processor 14 is additionally provided between a space filter processor and a driving processor. The signal processor 14 stores output values S and C of Sin and Cos filters which are phase-shifted by 90 degrees being detected at times k, k-1, and k+1 into a memory 15 (15a and 15b). Then, a denominator epsilon of an expression S0=Ck+1+Sk+1-Ck-1<2>-Sk-1<2>/2(Ck+1+Sk+1-k-1-Sk-1) (S0: amount Of deflection of filter output) is calculated 16 by using the output values S and C. the amount of deflection S0 is calculated 17 by this expression. Then, a criterion circuit 18 outputs an amount of deflection S0 (k-1) which is calculated 17 at the last time when the denominator epsilon is smaller than a value which is obtained from an experience and set 19 or an amount of deflection S0(k) which is newly calculated 17 in other cases to the driving processor.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a traveling speed and traveling distance detecting device for transfer cranes, traveling trucks, etc.

0002

BACKGROUND OF THE INVENTION For unmanned operation and automatic running of transfer cranes and traveling vehicles, it is essential to detect the traveling distance and traveling speed of the traveling device. Contact methods include methods such as determining the number of rotations of the wheels and methods such as determining the amount of wire drawn out. Non-contact methods include methods that use photoelectric converters to detect markers drawn on the road surface and methods that use lasers. etc.

[0003] Furthermore, a spatial filter method has been proposed that does not require drawing new markers or the like on the road surface. Particularly due to its durability and structural advantages, the adoption of non-contact methods has become mainstream, but in particular, the spatial filter method does not require any modification to equipment other than the traveling equipment, and it is possible to detect It is attracting attention as a simple detector because it can determine running speed and distance traveled.

The principle of the spatial filter method will be explained below with reference to FIGS. 3 to 9. Figure 3 shows a conceptual diagram when a spatial filter-type detection device is installed on a transfer crane, in which a television camera 8 is attached to the side of the crane's traveling wheels facing vertically downward so that the traveling road surface is in its field of view. , and a signal processing device (see FIG. 4) that spatially filters the video signal from the television camera to measure traveling speed and distance. In Figure 3, 1 is a container, 2
is the driver's cab, 3 is the spreader, 4 is the trolley, 5 is the chassis, 6 is the tires, and 7 is the girder. In FIG. 4, 9 is a camera amplifier, 10 is a pattern generator for specifying the spatial filter shape, 11 is a processor that performs spatial filter processing according to the pattern, and 12 is a travel speed and distance based on the output of the spatial filter processor. A traveling processor that calculates 1
3 is a control device for controlling the movement of the crane based on the results.

FIG. 5 is a television image 20 showing a driving road surface pattern imaged by a television camera 8, in which a lattice-like spatial filter pattern 21 is shown superimposed. Next, a spatial filter processing method will be explained. The most basic filter shape and effective for detecting travel is a grid. This lattice filter is installed at a certain pitch P perpendicular to the running direction of the screen. Further, the filters are arranged so that patterns having the same shape are shifted by 90 degrees (P/4). If these filters are distinguished by a and b, a will be a SIN filter and b will be a COS filter. Regarding the determination of pitch P and filter width W, etc.,
This needs to be done based on the measurement speed, accuracy, etc., but since this is not an essential problem, detailed explanation will be omitted. Spatial filter processing refers to the process of adding together all the luminance signals of images within a grid pattern. In this case, odd-numbered filters are weighted positively (+), and even-numbered filters are weighted negative (-). and add them together. This process is performed independently for the sin and cos filters. In addition, specifically, +1, 0, -1
This can be realized by superimposing the weighting pattern of the pattern generator 10 weighted according to the lattice pattern with the pattern photographed by the television camera and performing a sum-of-products operation.

[0006] In the above state, when the traveling device runs, the brightness pattern of the road surface moves from left to right in FIG. 5, so the above-mentioned spatial filter signal changes at a pitch of 2P. This situation is shown in FIG. Phase information and frequency information can be obtained from these sin and cos signals using the following equations.

[0007]

[Math 1]

[0008]

[Math 2]

[0009] Here, φk is the instantaneous phase at time k, fn is the instantaneous frequency, and Sk is the sin at that time.
The output value of the filter, Ck, indicates the output value of the cos filter. That is, the phase and frequency are determined using the values before and after a certain measurement time k. Among these, the frequency indicates the period in which the sin and cos outputs change,
The instantaneous traveling speed v is determined by the following equation.

0010

[Math 3]

Here, the filter pitch P can be determined as a length corresponding to the field of view of the image. On the other hand, the instantaneous phase is a signal that periodically changes every time the image moves 2P as shown in Figure 7, so by counting this signal, the instantaneous phase can be calculated using the following formula, where the number of counts is n. can.

0012

[Math 4]

[0013] Therefore, by calculating this X from a certain stop position, it is possible to detect the travel distance from there, that is, the moving position in the direction component perpendicular to the filter.

[0014]

[Problems to be Solved by the Invention] In the non-contact running speed and distance detection using the spatial filter described in the previous section, if the brightness pattern of the image captured by the television camera does not change while moving from left to right, then ( Math 1) and (Math 2)
The phase and frequency signals calculated by can be obtained normally. However, if the brightness changes, a bias So as shown in FIG. 8 will occur in the data in FIG. In particular, since the phase φk is determined from the ratio of changes in sin and cos as shown in FIG. 9, the influence of this error cannot be ignored in high-precision measurements. (Error = φk' - φk) Normally, the road surface condition does not change much when driving, so it does not change significantly, but the brightness pattern changes slightly due to sunlight and lighting, so the above effects can vary depending on the degree. This is a problem that will inevitably arise.

[0015] In order to utilize the principle of the spatial filter method particularly for objects such as cranes that are used outdoors and to put it into practical use, it is essential to solve the above problems. An object of the present invention is to provide a specific contact type traveling distance/speed measuring device for traveling equipment such as a transfer crane, which eliminates measurement errors caused by the spatial filter method due to changes in brightness due to sunlight or illumination. It is something.

[0016]

[Means for Solving the Problem] A value detected by the spatial filter method is corrected using bias So of the filter output. If the traveling gear is slow, the sine signal and co
Since the change in the s signal is small, when the rate of change in the filter output falls below a certain value, no correction is performed using the bias So.

[0017]

[Operation] In the spatial filter method, since the brightness pattern is used as a reference for speed/distance detection, errors will occur if the brightness changes due to changes in sunlight, illumination, etc. However, the error is reduced by correcting it using the bias So of the filter output. If the traveling speed becomes very slow and the changes in the sine and cosine signals become extremely small,
Since the correction based on the bias So increases the error, when the percentage change in the filter output becomes less than a predetermined value, the previous value is held.

[0018]

[Embodiment] An embodiment of the present invention will be explained with reference to FIGS. 1 and 2. In the block diagram shown in FIG. 1, component numbers 8 to 13 are similar to the prior art components shown in FIG.
4 is a newly added signal processor, and its specific calculation contents are shown in FIG. 15 is a memory that samples the sin and cos filter outputs and stores the values at times k-1, k, and k+1, and 16 uses those values to
A circuit 17 calculates the denominator ε of (Equation 6), which will be described later, is a circuit that calculates the bias amount So based on (Equation 6), and 1
8 is the previously calculated bias amount So if ε calculated by the calculator 16 is smaller than the value set by the setter 19.
(k-1), and if not, outputs the deviation amount So (k) newly calculated in step 17 to the running speed/distance calculator.

Note that the set value of the setter 19 is determined experimentally depending on the object to which it is applied. Now, if the bias of the filter output at a certain moment is So, then So
Assuming that does not change suddenly, the filter output is

[0020]

[Math 5]

It is expressed as: To this, sin2 θ+co
If we rearrange using the relational expression s2 θ=1,
(Sk − So )2 + (Ck − So )
2 = A2 Similarly, (Sk-1 - So )2 + (
Ck-1 −So )2 =A2 (
Sk+1 −So )2 +(Ck+1 −So )2
=A2 By eliminating A2 from the above three equations, the following two equations are obtained. Sk+12+Ck+12+Sk2-Ck2+2
(Sk +Ck -Sk-1 -Ck-1) So =
0Sk 2 +Ck 2 -Sk-12-Sk-12+
2(Sk-1 +Ck-1 -Sk-Ck)So
=0Of course, So can be calculated from each formula, but if we consider that the change is as large as possible and add both of them to calculate So,

[0022]

[Math 6]

##EQU1## is obtained, and the bias So at time k can be found using the values before and after time k. However, (k+
1) If the difference between time point and (k-1) time point is extremely small,
The denominator approaches zero, and the calculation error of So increases. To avoid this, it is possible to constantly monitor the denominator, and when it falls below a certain value (experimentally calculated), the calculation of So is not performed and the previous value is used as it is as the bias.

[0024]The reason for the small change is that the traveling speed has also decreased, and if the distance traveled is small, the change in bias during that time can be ignored, so as mentioned above, using the previous value It can be said that there is no problem in measurement. Based on So calculated above, So is subtracted from the biased filter output (Equation 5), and the unbiased sin and cos
Filter output can be calculated.

[0025] In this way, it is possible to calculate the traveling speed and distance by removing the influence of the bias So from (Equation 1) and (Equation 2).

[0026]

Effects of the Invention The present invention arranges grid-like spatial filters having phases different from each other by 90 degrees orthogonally to the traveling direction of a traveling road surface image captured by a television camera installed in a traveling device, and transmits the image signal to the spatial filter. In a device that non-contactly measures the traveling distance and traveling speed of the traveling device by performing ring processing, a calculation is performed based on the output values of two lattice-shaped spatial filters having a phase difference of 90 degrees. By correcting the brightness changes,
It has the following effects.

[0027] Highly accurate detection is achieved by correcting detection errors caused by changes in brightness of the road surface image, which are a problem especially when used outdoors, by correcting them using multiple values of the spatial filter output. can.

[Brief explanation of drawings]

FIG. 1 is a system configuration diagram of an embodiment of the present invention.

FIG. 2 is a block diagram of the signal processor of FIG. 1;

FIG. 3 is a schematic diagram of a container crane applying the spatial filter method.

FIG. 4 is a system configuration diagram of a conventional spatial filter method.

FIG. 5 is a drawing in which a grid-like spatial filter pattern is superimposed on the television image in FIG. 4;

FIG. 6 is a diagram showing the spatial filter signal during driving;

FIG. 7 is a diagram showing the instantaneous phase of a spatial filter signal.

FIG. 8 is a diagram showing bias in filter output.

FIG. 9 is a diagram showing errors in FIG. 8;

[Explanation of symbols]

8 TV camera 9 Camera amplifier 10 Pattern generator 11 Spatial filter processor 12 Travel processor 13 Travel control device 14 Signal processor

Claims (1)

[Claims] [Claim 1] Grid-shaped spatial filters having phases different from each other by 90 degrees are arranged orthogonally to the traveling direction of a traveling road surface image captured by a television camera installed in a traveling device, and the image signals are subjected to spatial filtering processing. Accordingly, in a device that non-contactly measures the travel distance and travel speed of the traveling device, brightness changes in the television camera image due to travel are calculated based on the output values of two lattice-shaped spatial filters having a phase difference of 90 degrees. A non-contact mileage/speed measurement device that corrects the following.