JPS5831626B2 - Eizou Shiyori Souchi - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a video processing device that converts a video image by processing the video imaged by an imaging device.

Conventionally, in order to differentiate, binarize, digitally transform, or apply a filter to a video, it has often been necessary to perform processing depending on the local situation of the video.

In other words, for example, when binarizing an image, there are of course cases in which it is acceptable to binarize the entire image using a constant threshold, a so-called fixed threshold, but sometimes this binarization threshold can be set to a floating value. In other words, when a point currently being scanned by an imaging device is binarized, it may be necessary to change the threshold for this binarization depending on the brightness in the vicinity of the point.

This is useful, for example, when multiple objects with different brightness enter the imaging device, and you want to convert the shape of each object into a neat binarized image with clear outlines.

Or, when you want to differentiate the image and emphasize the outline, you might apply uniform differentiation processing to the entire image.
A uniform contour line may not be obtained.

This is because, for example, the sensitivity of the imaging device is not uniform and changes depending on the brightness of the object being imaged.

Therefore, in such a case, a better contour image can be obtained by varying the differential gain, for example, depending on the average sharpness of a certain local image.

In this type of processing, the video signal obtained according to the scanning of the imaging device is one-dimensional, and therefore the video signal of a two-dimensional area cannot be obtained at the same time. The software first calculates the average brightness of a certain area, then determines a threshold based on that value, and uses that threshold to sequentially binarize the value of each pixel in the original image area. I had no choice but to take such measures.

An object of the present invention is to provide a concrete device that processes information of each pixel in a local video region according to the macro state of the region, and thereby enables various video processing. In particular, preprocessing in pattern recognition can be simplified.

In order to achieve the above object, the video processing device of the present invention divides the clear image signal into two paths, one path temporarily stores the video of a certain screen area, and the other path stores the video of the same screen area. Configure it to process video.

This processing corresponds to calculating the average value of the image in that screen area or counting the number of points where the brightness changes suddenly, and is called a macro information detection circuit in the sense that it is a process for understanding the situation in that local area. I'll call you.

In this macro information detection circuit, when the imaging device finishes scanning the area of interest, the macro information for that area is completed and the result is output.

Furthermore, this point in time is exactly the point in time when this local video information has finished being stored in, for example, a shift register type primary storage device.

Therefore, the primary storage circuit, which is the first path, sequentially reads out the video information delayed by that area, and processes it into information for each pixel one after another using the macro information already detected in the second path. can be added.

Hereinafter, the present invention will be explained in detail with reference to the drawings.

FIG. 1 shows an embodiment of the present invention.

A video signal 2 from the imaging device 1 is converted into a digital quantity by an AD conversion circuit 3 and guided to a temporary storage circuit 6 and a macro information detection circuit 7 via gates 4 and 5.

Signal A for opening gates 4 and 5.

yBO is, for example, a signal that logically becomes l'' during a time period when the imaging device is scanning a certain two-dimensional small area of interest, and this will be described later.

The temporary storage circuit 6 has a storage capacity sufficient to store information from this two-dimensional small area.

Normally, it is desirable to design this storage capacity to be the same as the amount of information generated in the two-dimensional small area, and to make the two-dimensional small area the unit size for video processing.

Information input through the gate 5 is processed by a macro information detection circuit 7, and the result is held in an information holding circuit 8, such as a sample and hold circuit or a register.

The timing of this holding is the signal C8 at the time when the scanning of the two-dimensional small area currently being focused on is completed, that is, the time when the processing in the macro information detection circuit 7 is completed. is held by 8, and the macro information processing circuit 7 is reset to return to the initial state to prepare for processing in the two-dimensional area to be processed next.

That is, when the macro information is held in the information holding circuit 8, the game 14 is used again for the next two-dimensional area.
, 5 is opened, and the macro information processing circuit 7 calculates the macro information in the new two-dimensional area using the pixel information that is input one after another, and the video information in this new two-dimensional area is stored in the temporary storage circuit. 6 is stored.

This temporary storage circuit 6 is, for example, a shift register type,
It is easy to configure the system so that every time information in a new two-dimensional area is input, information in the previous two-dimensional area is pushed out.

That is, at this time, the temporary storage circuit 6 can be regarded as a delay circuit for information outside the two-dimensional area, which is a unit of video processing.

Therefore, the information 9 output from the temporary storage circuit 6 may be processed by the processing circuit 11 using the information 10 already obtained and held in the holding circuit 8. At this time, the output 12 of the processing circuit 11 is This is a converted video that takes into consideration the situation.

In the above explanation, 3 is an AD converter, but this is not necessarily necessary, and it is possible to process analog quantities as they are by selecting an appropriate primary storage circuit 6 and macro information detection circuit 7. Not even.

3 in this case may generally be regarded as a preprocessing circuit.

FIG. 2 shows the control signal A in FIG.

,Bo. 2 is a diagram for explaining Co, and shows the image plane of the imaging device 1. FIG.

In FIG. 2, it is explained that two points are provided in the horizontal scanning direction of the image plane (screen) and a logic signal A is generated having the coordinates Xo and X1 of these points as the starting and ending points, respectively.

Similarly, the starting point Y in the vertical scanning direction.

, it is possible to set two coordinates, the end point ¥1, and create a signal B that is logically "1" between these two points, and the end point ¥1.
It is also easy to create a logic signal C8 that becomes "1" when .

The signal corresponding to the logical product of these A and B signals corresponds to the signal in the small area indicated by the needle line in Fig. 2, so this is
A in the diagram.

, can be used as a BO. The ability to create these signals will be discussed later.

How to generate these signals will be described later.

In this case, by changing the coordinate values of Xo, xl, Yo, and Yl, the small area indicated by diagonal lines can be moved to an arbitrary position on the screen.

FIG. 3 shows a control signal A of the apparatus of the present invention shown in FIG. 1 based on the principle of FIG. 2.

, Bo, and co.

The four coordinates X mentioned above.

, Xl, Yo, ¥1 are registers 20, 21゜22.2
3 respectively.

Now, the picture element pulses P from the pulse oscillator 24 are counted by an It is composed of

This X synchronization signal 26 is counted by a Y counter 27 that counts the number of scanning lines for -vertical scanning.
Configure to reset.

At this time, horizontal and vertical synchronization signals 26.28 are obtained by the pulse P corresponding to the picture element from the pulse oscillator 24,
These are used as scanning synchronization signals for the imaging device 1 shown in FIG.

When this is done, the contents of the X counter 25 and Y counter 27 in FIG. 3 correspond to the horizontal and vertical positions of the screen at the moment when the imaging device is currently scanning. , 22゜ 23 contents to match circuit 30
, 31, 32, and 33, and when a match is found, the signals are led to two flip-flop circuits 34 and 35.

That is, if the flip-flop circuit 34 is set when there is a match with Xo, and the flip-flop circuit 34 is reset when there is a match with X, the output from 34 will already be the signal A shown in FIG. .

If the flip-flop circuit 35 is set when it matches Yo and reset when it matches ¥1, the signal B shown in FIG. 2 will be obtained as its output.

Therefore, if the signal Bo is generated by the AND circuit 36, this is a signal that becomes 1" while scanning the aforementioned small area, and the first
It can be used as an opening/closing signal for the gate 5 shown in the figure.

Also, the picture element pulse signal P is this B.

If the signal is gated by the gate circuit 37, the signal A is generated.

is obtained, and this signal becomes a pixel pulse only when scanning the small area in FIG. 2, and is used as the opening/closing signal of the gate 4 in FIG. can.

AOt BO may be used for either gate 4 or 5 as appropriate.

Furthermore, if the output of the flip-flop circuit 35 is differentiated by the differentiating circuit 38 to create a signal C6 when 35 is reset by Y, this is already a signal indicating that scanning of a small area has been completed, as shown in FIG. This signal is used as a timing signal for holding information in the information holding circuit 8 shown in FIG.

In this way, if the contents of the registers 20, 21, 22, and 23 are changed one after another, the screen of the small area changes accordingly, and the process progresses.

At this time, the area of the small area is the temporary storage circuit 6 shown in FIG.
It is convenient to design the size so that it matches the size of xo, Xl,
There is no longer a need to specify Y, , and Y.

That is, if the width of the small area N XI Xos Yl-Yo is fixed in advance, it is easy to create a circuit configuration in which only two coordinates, Xo and Yo, need to be changed.

To specify xo, Yo, etc., for example, a command from a computer may be used, or a control circuit may be provided to sequentially update these values.

FIG. 4 is a diagram for explaining a method of applying processing to the entire image by sequentially switching small areas.

In the figure, the screen is divided vertically and horizontally, each forming a small area of the video.

When scanning the first image field, the leftmost X force direction 0th
The th vertically long strip-shaped area becomes the processing target, and at this time, as the vertical scanning progresses in the Y direction, the 0th, 1st, and n-th
The first small area is selected as the processing area.

When the image field changes, the first vertical strip-shaped area in the X force direction becomes the target of processing, and similarly, the 01st 1st, ..., n-1st, etc. in the Y direction, and so on. Processing proceeds in order from the top.

The figure just shows the situation when the second band-shaped region in the X force direction is selected, and at this time, the diagonally shaded capitula group is selected as the processing target.

In other words, all you need to do is create a logic signal like A in the same figure that is synchronized with the horizontal scanning signal.
All you have to do is shift the case.

Such a sequential small area selection method can be realized as shown in FIG. 5 below.

FIG. 5 shows a circuit for sequentially switching small areas and applying processing to the entire surface based on the method shown in FIG. 4, and shows a control signal A of the apparatus of the present invention shown in FIG.

. This is another example of a specific circuit configuration for obtaining Bo and Co.

The picture element pulse P from the pulse oscillator 24 is sent to the X counter 2
5, and when it finishes counting the number of picture elements in the horizontal scanning direction, it outputs an X synchronization signal 26 and resets its own counter 25.

When this signal 26 is counted by a predetermined vertical scanning line by an X counter 27, it outputs a Y synchronization signal 28 and resets the counter 27.

This X, Y synchronization signal 26°28 is used as a drive signal to the imaging device 1 in the apparatus of the present invention shown in FIG.

This Y synchronization signal 28 is input to a field number counter 40 and counted.

This field number counter 40 is configured to issue a processing end signal 41 and reset itself when a count equal to the number m of small area divisions in the horizontal direction of the screen shown in FIG. 4 is made.

Therefore, the contents of the counter 40 are such that the first field is 01, the first field is i, -..----1 field is ml, and so on.

First, the picture element pulse P is guided to the X-direction region width counter 42, and when this counter counts a number corresponding to the width of the small region, it resets itself and adds 1 to the X-direction region number rank 43.

In this way, the contents of the counter 43 will be incremented by 1 each time the area changes in the scanning in the X direction in FIG. , . . . , the m-th strip region becomes ml.

Therefore, if the minus number circuit 44 generates a signal A that becomes "1" when the contents of the two counters 40 and 41 match, then in the first field, when scanning the leftmost band-shaped area in FIG. ”, and in the first field, when scanning the second strip area from the left, it becomes “1”,
As the fields progress, the position of 1'' shifts to the right, and the entire screen is processed by scanning m fields.

This signal A is the signal B in FIG.

It is used by closing it. Further, the signal A is obtained by gating the pixel pulse P by the gate circuit 45 based on this signal.

used as. On the other hand, the X synchronization signal 26 enters the Y-direction area width counter 46, and when the number of scanning lines corresponding to the Y-direction area width is counted, the signal C8 is output and the counter 46 is reset by itself.

In this way, the signal becomes a signal as shown in FIG. 4, which is used as co in FIG.

These A. , Bo, and co control the apparatus of the present invention in FIG. 1, the processing from the upper left to the bottom in FIG. 4 will be sequentially applied from the left to the right as the frame progresses.

FIG. 6 shows a specific example of the structure of the macro information detection circuit 7 shown in FIG.

In the figure, a is a combination of an integrator and an attenuator, and when brightness information is input, it is integrated over one small area.

In other words, the average spacing within the small area is determined.

When scanning of this small area is completed, the output is led to a holding circuit, and the integrator capacitor is short-circuited to return to the initial state in preparation for processing in the next small area.

Similarly, figure b is an example of counting the number of sudden change points in the brightness of the image within a small area, 51 is a differentiation circuit, 52 is an absolute value circuit, and from here the degree of sudden change in the brightness of the image is counted. A large positive output can be obtained.

Therefore, a threshold circuit 53 is provided which outputs "1" when this output exceeds the threshold value .theta., and the output thereof is counted by a counter 54.

The contents of this counter can be transferred to a register as a holding circuit at the end of area scanning, and the counter can be configured to be reset.

FIG. 7 shows a specific example of the structure of the temporary storage circuit 6 shown in FIG.

Several shift registers 60 are connected in series so that the length of one shift register corresponds to the width of the small area in the X direction, and the number of shift registers corresponds to the width of the small area in the Y direction.

In this way, the output 9 lags behind the input by one small area.

FIG. 8 shows a specific configuration example of the processing circuit 11 shown in FIG.

a is used when subtracting the average brightness of a small area from each pixel information.

A shown in the figure means the output from a in FIG. 7.

62 is an adder. Figure b shows the process of extracting only the waveform that exceeds the average value.

If the subtraction result 63 between signals 9 and 10 by the adder 62 is positive, the output of the threshold circuit 64 becomes "1",
passing the signal 63 through an analog gate circuit 65;
If the output of 62 is negative, the output of 64 will be O”,
Analog gate circuit diagram C is an example of calculating the average spatially and subtracting the average value of the area, and a, b, c, and d in the diagram indicate the reading positions from the temporary storage circuit in FIG. 7. ing.

As described above, the macro information detection circuit 7, the processing circuit 11, etc. according to the present invention can have various configurations.

FIG. 9 shows another configuration example of the present invention, in which the configuration of FIG. 1 is connected in multiple stages.

The processed video signal 12 outputted from the processing circuit 11 passes through the gate 4' again and enters the temporary storage circuit 6', and also passes through the gate 5' and enters the area macro information detection circuit 7'.

When scanning of the area is completed, the macro information transferred to the holding circuit 8' is used to process the video information from the processing circuit 11' again and output as i, 12'.

The video signal 12' after this processing enters the next stage again.

By configuring the images in multiple stages in this manner, the images can be transformed one after another.

FIG. 10 shows another configuration example of the present invention, which enables the multi-stage processing shown in FIG. 9 with a single-stage configuration.

The video signal 2 from the imaging device 1 passes through the preprocessing circuit 3, is guided to the gates 4 and 5 via the switch 71, and is sent to the temporary storage circuit 6.
and the macro information detection circuit 7.

The macro information 10 detected in step 7 and held in holding circuit 8 is processed together with output 9 from temporary storage circuit 6 in processing device 11 to output a modified video signal 12.

This video signal 12 is guided to a switching circuit 71.

Therefore, if this switch is switched in response to a command from the arithmetic and control unit 70 such as a computer, then 1
2 enters 6 and 7 via 4 and 5.

At this time, the processing contents of the macro information detection circuit 7 and the processing circuit 11 are appropriately switched in response to a command from the arithmetic control unit 70, and processes with different contents are successively added to the video information of the same area to produce a desired video signal. It can be converted.

If processing that does not require using the macro information of the area is required in this multi-stage processing process, it is possible to control the macro information detection circuit 7 to cut off this path so that only the information from 6 enters 11. Of course it can be done.

The signal converted into the desired image is one in which characteristic parts such as outlines are emphasized, and is led to separate dedicated hardware or a general-purpose computer to recognize the object's shape, position, name, etc. be able to.

In the above description, the Rask scanning force type imaging device used in ordinary TVs has been described.

In this case, we have described a method in which areas are selected sequentially and a method in which areas are selected at random based on instructions from a computer, for example.

However, if the imaging device itself is, for example, an imaging sector tube or a flying spot tube, it goes without saying that scanning itself can be limited to a small area and the scanning areas can be sequentially switched.

However, accurately controlling the scanning area with such devices is generally difficult or requires complex control circuitry.

According to the present invention, since a rusk scan type imaging device is used, it is possible to specify a local area only by timing.
The device is simple.

According to the invention described above, it is possible to process images using relatively small-scale hardware, and it is also possible to perform processing that takes into account macro information of an area.

This device can be applied to preprocessing devices for pattern recognition devices, as well as industrial vision devices for the purpose of sorting and inspecting objects in production processes.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing one embodiment of the present invention. FIG. 2 is a diagram showing a screen for explaining the control signal in FIG. 1. FIG. 3 is a block diagram showing a control signal based on FIG. 2. 4 is a diagram showing a screen for explaining another example of the configuration for obtaining the control signal in FIG. 1,
FIG. 5 is another specific configuration example for obtaining a control signal based on FIG. 4, FIG. 6 is a specific configuration example of a macro information detection circuit which is a main component of the device of the present invention, and FIG. 8 is a specific example of the structure of a temporary storage circuit, FIG. 8 is a specific example of a processing circuit, FIG. 9 is another example of the structure of the present invention, and FIG. 10 is yet another example of the structure of the present invention. . 1... Imaging device, 2... Video signal, 3
...AD converter, pre-processing circuit, 4...
Gate, 5... Gate, 6... Temporary storage circuit, 7... Macro information detection circuit, 8...
...Holding circuit, 11... Processing circuit, 20-2
3...Register, 24...Pulse oscillator, 25...X counter, 27...Y
Counter, 30-33... Matching circuit, 34, 3
5...Flip-flop circuit, 36...
・Logic product circuit, 37... Gate circuit, 38...
... Differentiation circuit, 40 ... Field number counter, 42 ... X-direction area width counter, 43.
...X-direction area number counter, 44... Match circuit, 45... Gate circuit, 46...
・Y-direction area width counter, 51...differentiation circuit,
52... Absolute value circuit, 53... Threshold circuit, 54... Counter, 60...
Shift register, 62... Adder, 64...
... Threshold circuit, 65 ... Analog gate circuit, 70 ... Arithmetic control device, 71 ...
・Switcher.

Claims (1)

[Claims] 1. The scanning area of the raster scanning type imaging means is divided into a plurality of rows of regions in the horizontal scanning direction, and each row is further divided into a plurality of two-dimensional local regions in the vertical scanning direction. An apparatus that performs image processing over the entire scanning area by performing image processing on an area-by-area basis, wherein (a) among the outputs of the imaging means, a video signal belonging to one row area is sequentially generated for each vertical scanning period; (b) a shift register-like storage means into which the selected video signal portions are sequentially input and which has a capacity to store the video signal portions in each two-dimensional local region; (c) a means for selecting only the video signal portions; The selected video signal parts are input sequentially,
means for detecting a macro situation of one two-dimensional local area each time input of a video signal portion in one two-dimensional local area in each column area is completed; (d) the detection includes the result of the detection; and (e) means for converting video signal portions sequentially output from the shift register storage means into other signals based on the stored detection results. 2. (a) means for selecting only a video signal portion belonging to a two-dimensional local region within the scanning area from among the output of the raster scanning type imaging means; and (b) means for sequentially inputting the selected video signal portion and a shift register-like storage means having a capacity to store all the video signal portions in the local area; (c) the selected video signal portions are sequentially input, and upon completion of input of all the video signal portions in the local area; (d) means for storing the detection result in synchronization with the detection; and (e) the shift register-like storage based on the stored detection result. (f) means for converting the video signal portions sequentially output from the means into other signals;
The selection means is a means for selecting the converted signal after the selection of all video signal portions belonging to the local area is completed. 3. The second detecting means is configured to detect different macro information when the selecting means selects the video signal portion of the local area and when the converted signal is selected. Video processing device. 4. The video processing device according to item 2, wherein the conversion means performs different signal conversions when the selection means selects the video signal portion of the local area and when the converted signal is selected.