JPS61167293A - Method for processing video signals - Google Patents

Abstract

PURPOSE:To reduce the memory capacity and data processing time by storing picture image scanning signals binary coded in light and dark after compression into data concerning the respective time widths of light and dark periods and performing required processing on the basis of these storted data. CONSTITUTION:Line image sensors 1a, 1b pick up images of paper sheets 2, 2' respectively having apertures 2a, 2a,. A circuit 4 selects either the control signal from a switching control circuit 3 or clocks CLK from an oscillating circuit 5 to supply the resulting output to a threshold determining circuit 7. The circuit 6 compares the picture image scanning signal and the threshold value from the circuit 7 to provide a binary signal which is inputted to a latch circuit 8 and to a rise-fall detecting circuit 10. The circuit 10 generates pulses at the changing time of the binary signals to supply latch instruction to a circuit 8 while supplying them to a counter 12. The circuit 8 latches the values of CLK by a counter 11 by time widths of light and dark periods to store them with the count values of the counter 12 as addresses in a memory 14 while storing them in a main memory 17 through CPU15 by the address of the counter 12.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for processing image signals used in image recognition and the like.

[Prior art]

When using a computer to recognize images captured by a camera such as a blurry camera, the captured images are temporarily stored in a video RAM (continuously reloading memory 9), and then the stored image information is read out and processed as required. We are planning to do this.

[Problem that the invention seeks to solve]

However, this method requires a video RAM with a capacity corresponding to the number of pixels, and since the stored information is for each pixel, there is a problem that it takes a long time to process the data read from the video RAM. Oh, it was. This also applies to one-dimensional imaging devices.

[Means for solving problems]

The present invention has been made in order to solve such problems, and the present invention compresses and stores an image scanning signal that has been binarized into bright and dark periods into data relating to a width of 9 hours each for bright and dark periods, and stores this stored data. An object of the present invention is to provide an image signal processing method that can reduce memory capacity or data processing time by performing necessary data processing based on the following.

〔Example〕

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to drawings showing an embodiment in which the present invention is applied to a line image sensor, that is, a carving recognition device using a one-dimensional imaging device.

In Fig. 1, 1 and 1' are line image sensors I, respectively.
An image pickup device consisting of a combination of a, lb and a lens (not shown), which is a line image sensor la.

1b is arranged at a position separated from the lens by an appropriate distance. The object to be photographed is, for example, a horizontally long rectangular piece of white paper 2°2' with holes 2a and 2'a made in the center, and the background is dark.

The switching control circuit 3 supplies a signal for selecting the line image sensors Ia and lb to the input circuit 4, as shown in FIG. 2(A). The oscillation circuit 5 also provides the input circuit 4 with a clock signal CIJ shown in FIG. 2(d).

The input circuit 4 issues an enable signal to the line image sensor 1a or 1b to be operated according to the signal from the switching control circuit 3. The line image sensors la and Ib have a large number of photosensitive elements arranged in parallel in one row, and when an enable signal is applied and a clock signal CLK is applied via the input circuit 4, the line image sensor 1a or 11] is scanned, and an electrical signal of a level corresponding to the light receiving bar of each photosensitive element, that is, an image scanning signal, is output in synchronization with each pulse of the clock signal C1.

This image scanning signal is applied to a comparison circuit 6 and a dark value determination circuit 7. The dark value determination circuit 7 is a circuit that determines the dark value level of the comparison circuit 6 by using a specific signal portion such as the peak value of the image scanning signal. The comparison circuit 6 uses this dark value level to binarize the input image scanning signal. Figure 2 (
C) shows this binarized signal.

The binarized image scanning signal is applied to a latch circuit 8 and a rise/fall detection circuit 10. Rising/falling detection circuit 10
is constructed using a differentiating circuit or the like, and cuts out a pulse signal with a short time width at the rising and falling points of the image scanning signal. This pulse signal is given to the launch circuit 8 as a signal instructing launch, and is also given to the counter 12 as a counting object.

The switching control circuit 3 has a switching signal as shown in Fig. 2 (b) in addition to the switching signal to be selected.
Line image sensor la as shown.

An operation command signal of lb is issued, and when this is at the "H" level, an enable signal is issued in synchronization with this. The operation command signal is also provided to the counters 11, 12 as a reset state. The counters 11 and 12 are in a reset state when the operation command signal is "L" and count the stitches when it is "H".

A clock signal CIJ is applied to the counter 11 as a counting target. The counted contents of the counter 11 are given to the launch hI circuit 8, and the counted contents of the counter 12 are given to the latch circuit 9 and the changeover switch 2. The latch circuit 8 is also given an enable signal output from the switching control circuit 3. Therefore, the latch circuit 8 determines the count contents of the counter 11 when the binarized image scanning signal changes from bright to dark or from dark to bright, which line image sensor Ia or Ib is outputting at that time, and the image concerned. Latch the brightness and darkness of the scanning signal (immediately after the change).

The operation command signal output from the switching control circuit 3 is sent to the launch circuit 9.
However, the latch circuit 9 launches the stitch number content of the counter 12 at the falling timing.

The latched contents of the latch circuit 8 are given as write data to the memo 1J14 via the changeover switch 22 which is controlled by the operation command signal. When the contents of the memory 14 are to be read out, the changeover switch 22 is switched to the side of the data bus 15 connected to the central processing unit 13 so that the contents of the memory 14 can be transferred to the central processing unit 13.

The data latched in the latch circuit 9 is also transferred to the central processing unit 13 via the data bus 15. Also counter 1
The number of stitches 2 is given to the memory 14 as a write address signal via a changeover switch 21 which is controlled by the central processing unit 13. Further, an address signal applied from the central processing bag W13 via the address bus 16 is read out to the memo 1J14 via the changeover switch 21 and applied as a read address signal. The changeover switches 21 and 22 switch the counter 12 and launch circuit 8 while the operation command signal is at the "H" level.
while the address bus 1 is switched to the “L” level.
5. It is switched to the data bus 16 side. Central processing bag M1
The data successively outputted from the memory 14 according to the command No. 3 is transferred and stored in the main memory 17, and similarly processed.

In addition, although not particularly shown, input means for the central processing unit 13, means for displaying input contents or calculation contents, etc. are provided.

Next, the method of the present invention executed by the above-mentioned apparatus will be explained. As shown in FIG. 2(b), when the operation command signal reaches the "■" level, the switching control circuit 3 outputs an enable signal for the line image sensor 1a. Counters 11 and 12 also start counting stitches.

FIG. 2(c) shows an enlarged view of the period in which the operation command signal in FIG. 2(b) is "H level". 2. Since it is the background, the binary image signal is dark, bright, dark, bright as shown in FIG. 2 (c).

It becomes dark.

Now, when the binary image signal changes from dark to bright, the launch circuit 8 launches using the pulse signal output by the rise/fall detection circuit 10.

The information latched at this time is the rising edge of the operation command signal (I2
Counter 1 corresponds to the time from the start of scanning from one end of the line image sensor 1a to the time of the change from dark to bright, and therefore the length from the edge of the captured image to the edge of the paper.
Immediately after the needle value of 1 changes by 01, there is a difference between light and dark, that is, brightness, and line image sensor la and lbO, that is, 1a. On the other hand, the counter 12 is incremented by one step by the pulse signal from the rise/fall detection circuit 10. The contents latched by the latch circuit 8 are written into the memory 14 using the contents of the counter 12 as a write address.

Thereafter, each time the binary image scanning signal changes from bright to dark or from dark to bright, writing to the latch and memo 4 is performed as described above.

Table 1 is a list of the contents lunched in this way and written in the memo January 4.

When the operation command signal falls after writing to the memory 4 is performed as shown in Table 1, the latch circuit 9 latches the count 4 of the counter 12 at that time.

The latch circuit 9 stores the final address at which the data in the memory 14 is written.

Then, when the operation command signal becomes "L" level, the counters 11.12 are reset, while the changeover switches 21.22 are turned on the address bus 16.12. Switch to the data bus 15 side. The central processing unit 13 reads the contents of the latch circuit 9, learns that the final address is 4, sequentially issues addresses 1, 2, 3.4 of the memory 14, reads the contents, and stores them in the main memory 17. do.

Next, when the operation command signal goes to "H" level, similar processing is performed on both images captured by the line image sensor lh.

In this way, the image information captured by the line image sensors Ia and lb is transferred from the memory 14 to the main memory 17,
will be stored here.

Note that although the above-described embodiments are those in which a plurality of one-dimensional imaging devices are used, the present invention can also be applied to one that uses one or a two-dimensional imaging device. Even in the case of a two-dimensional imaging device, the contents of the memory 4-hernch circuit 8 may be written for each signal of the 1 scanning line.

Further, in the above embodiment, the latch circuit 8 latches the enable signal output from the switching control circuit 3, but the difference between the line image sensors la and lb is 1.
Since it does not change between two operation command signals, the launch circuit 9 may be configured to store it once for each operation command signal.

Note that in the case of a two-dimensional imaging device, line numbers are to be stored.

Furthermore, the number of stitches on the counter 11 is not counted from the time when the operation command signal reaches 1, but is reset by the pulse signal output from the rise/fall detection circuit 10, and the time width of darkness and light is directly counted. It may also be a thing.

Furthermore, the portion to be counted by the counter 1 may not be the entire imaging field of view, but may be limited to a portion thereof.

〔effect〕

As described above, in the present invention, since the information contained in the binarized image scanning signal is compressed and stored, the memory capacity can be extremely small compared to the case where data is stored for each pixel. In addition, this saves not only writing and successive output times but also data processing time, making it possible to shorten the calculation time of the control system as a whole.

In the above embodiment, the central processing unit is responsible for the operation of the memory (SC) from Memo January 4 to Main Memo 1.
Since the data is only used during the transfer and storage to the month 7, the utilization efficiency of the central processing unit can be increased, and the calculation time for shape recognition, measurement, etc. can be shortened.

Furthermore, since the final address written in the memo January 4 is stored in the latch circuit 9, there is no waste of transferring the entire capacity of the memory 14 to the main memory I7, and the present invention has excellent effects. .

[Brief explanation of the drawing]

FIG. 1 is a schematic block diagram showing the implementation state of the present invention, and FIG. 2 is a time chart for explaining the method of the present invention.

Claims (1)

[Claims] 1. The image scanning signal that has been binarized into bright and dark is compressed and stored into data regarding the time width of each of the bright and dark periods, and the required data processing is performed based on this stored data. How to process image signals. 2. Compressing the image scanning signal that has been binarized into light and dark into data regarding the time width of each of the light and dark periods and storing it in a first memory, and also storing the final address of the compressed and stored data in the first memory. . A method for processing an image signal, characterized in that the stored contents of the first memory up to the final address are transferred and stored in a second memory, and required data processing is performed based on the stored data.