JPH0113149B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention detects the position of a specific point (for example, a contour) of an image, that is, the position of the intersection of the image signal from the image sensor and the threshold value (which should correspond to the specific point). , relates to an image position detection method in which the image position is determined by linear interpolation from the video information possessed by the pixel.

An image sensor is a device in which photoelectric elements (pixels) with a microscopic area are arranged at a fixed pitch, the illuminance detected by each pixel is exchanged into an amount of electricity, and the data held by each pixel is sent out in series.

Now, for example, as shown in the relationship diagram between each pixel output and the threshold value in Fig. 1, when light is irradiated along the scanning direction SCN of the pixels of the image sensor, each pixel... -1, n, . . . detect the average illuminance on the light-receiving surface, so the output of the image sensor becomes a step-like video signal V.

Here, the threshold value that does not match the output of any pixel
When SH is set and an attempt is made to find the position of an image that intersects this threshold, one of the adjacent pixels n and n-1 that have outputs P and Q whose magnitude relationship is opposite to the threshold SH Although it is simple and easy to find the position of the pixel, this method has a problem in detection accuracy in that the position of the pattern can only be resolved into dimensions corresponding to the pixel pitch.

As a means to solve this problem, changes in light between pixels, which cannot be obtained directly from the image sensor, are obtained by linear interpolation of the image information of adjacent pixels, and virtual pixels formed on a straight line are There is a method of finding the intersection point with the threshold value.

However, the image position detection method that detects the intersection point by linear interpolation is based on the following off-line processing, which requires a long time and is not economical.

(1) Convert the video information of all pixels output from the image sensor into digital values and temporarily store them in the memory circuit.

(2) Through digital comparison processing, search for adjacent pixels whose outputs are opposite in magnitude with respect to a desired threshold value.

(3) Linear interpolation is performed between the outputs of the pixels obtained in this way, and the intersection with the threshold value is found.

The purpose of the present invention is to eliminate the above-mentioned drawbacks of the prior art, improve the detection resolution of image positions, and
An object of the present invention is to provide an economical image position detection method that speeds up detection processing.

The feature of the present invention is to perform analog/digital conversion of a serial video signal obtained from an image sensor,
The digital values corresponding to the two adjacent pixels are sequentially stored and shifted, and the magnitude relationship between these digital values and a predetermined threshold value is compared.
When the digital value corresponding to the pixel reaches the intermediate value, by performing linear interpolation calculation processing according to the count result of the transfer clock of the serial video signal,
This image position detection method determines the intersection between the digital values corresponding to the two pixels and the threshold value, and uses the result as a detection output of a desired image position.

In short, position detection processing is performed online in accordance with the transfer clock of the serial video signal, which is generally high speed, to make it faster and more economical.
The aim is to improve its resolution.

Embodiments of the present invention will be described below based on the drawings. FIG. 2 is a system configuration diagram of an embodiment of the image position detection method according to the present invention, and FIG. 3 is a time chart thereof.

Here, 1 is an image sensor, 2 is a clock generator, and 3 is an A/D (analog/digital)
Converters 4 and 5 are first and second timing circuits, respectively; 6 and 7 are first and second registers, respectively; 8 is a delay circuit; 9 and 10 are first and second timing circuits, respectively; 2nd comparison circuit, 11 is an exclusive OR circuit, 12 is an AND circuit, 13, 14, 1
5 are first, second, and third storage circuits, respectively;
16 is a counter, 17 is a threshold value setting circuit, 18
is an arithmetic circuit, and 19 is an intersection output terminal. It should be noted that signals a to m in FIG. 3 are the same as those indicated by the same symbols in FIG. 1.

First, the image sensor 1 receives information on the illuminance detected by each pixel upon receiving light from the image to be detected.
In synchronization with clock a from clock generator 2,
Outputs a serial video signal c. Since each pixel detects the average illuminance on the light-receiving surface, the serial video signal c is synchronized with the clock a as shown in FIG.
Each wave appears as a staircase wave.

The counter 16 detects the position d of the currently output pixel by counting the clock a. On the other hand, the first timing circuit 4 receives the clock a, generates a timing pulse, and gives the following commands to each required circuit.

(1) Instruct the A/D converter 3 to start conversion for each pixel output.

(2) After one scan is completed (after all pixel output is completed), clear the contents of the counter 16.

(3) Strobe the outputs of the comparison circuits 9 and 10 for each pixel output.

When the A/D converter 3 completes A/D conversion for each pixel, it first transfers the information of the previous pixel stored in the first register 6 to the second register 7 in response to a conversion end command. At the same time, the immediately following conversion information is written into the first register 6. A shift operation to the second register 7 and a shift operation to the first register 7
The time lag in writing to the register 6 is achieved by delaying the conversion completion signal in the delay circuit 8.

Here, the video information (A/D
(after conversion) is expressed as V(n), the first register 6 and second register 7 always hold information V(n) and V(n-1) of adjacent pixels, respectively. It turns out.

Next, the information V(n) and V(n-1) of adjacent pixels held in both registers 6 and 7 are respectively
The first comparison circuit 9 and the second comparison circuit 10 compare the magnitude relationship with the threshold value b. threshold b
is set in advance in the threshold value setting circuit 17 and is applied to both comparison circuits 9 and 10. The outputs of these comparison circuits 9 and 10 indicate that the value on the threshold input side is A,
Assuming that the value on the video information input side is B, the following logic is output depending on the magnitude relationship between A and B.

“1” (or “0”) when A≧B, A<B
“0” (or “1”) when .

Therefore, if the value of the threshold b is T, the conditions for the output h of the exclusive OR circuit 11 which receives the outputs of both the comparison circuits 9 and 10 to be "1" are as follows: V(n)≧T, and , V(n-1)<T or V(n)<T and V(n-1)≧T.

In the above relational expression, the threshold value b is the video information V(n), V
(n-1).

Under these conditions, exclusive OR circuit 11
When the AND circuit 12 is opened by the timing pulse i from the second timing circuit 5, the output h of the first, second, and third memory circuits 13,
This is a write command for 14 and 15. That is, for example, the first storage circuit 13 stores video information V.
(n), the second storage circuit 14 stores the same V(n-1), and the third storage circuit 15 stores the pixel position n.

The arithmetic circuit 18 uses these three types of information j, k, l
Then, the calculation process AL of linear interpolation regarding the information of adjacent pixels is performed as shown in the following equation, and the intersection point t with the threshold value b is determined and outputted to the output terminal 19.

t=n-{V(n)-T}/{V(n)-V(n-
1)} The second term on the right side above indicates that adjacent pixels sandwiching the threshold are divided by proportional allocation, which has the same effect as increasing the number of pixels (reducing the pixel pitch). is obtained.

As described in detail above, according to the present invention, the intersection point between a threshold value and a video signal can be found by linear interpolation between adjacent pixels sandwiching the threshold value, and the intersection point can be obtained at the same time as the video signal output. Since the above processing can be performed on the axis (online),
A high-speed, high-resolution image position detection method can be realized.
Furthermore, economy can be achieved by reducing storage devices and simplifying processing contents, and the effects are remarkable.

[Brief explanation of drawings]

FIG. 1 is a diagram showing an example of the relationship between each pixel output and a threshold value, FIG. 2 is a system configuration diagram of an embodiment of the image position detection method according to the present invention, and FIG. 3 is a time chart thereof. 1... Image sensor, 2... Clock generator, 3
...A/D converter, 4, 5...timing circuit, 6,
7... Register, 8... Delay circuit, 9, 10... Comparison circuit, 11... Exclusive OR circuit, 12... AND circuit, 13, 14, 15... Memory circuit, 16... Counter, 17... Threshold value setting circuit, 18... Operation circuit, 19
...Intersection output terminal.

Claims (1)

[Claims] 1 converting the serial video signal obtained from the image sensor from analog to digital, and calculating the pixel position by counting the transfer clock signal for outputting the serial video signal from the image sensor,
The digital values (V(n-1), V(n)) of the above-mentioned serial video signals corresponding to two adjacent pixels are sequentially stored.
and their digital values (V(n-1),
The magnitude relationship between V(n)) and a predetermined threshold (T) is compared, and the threshold (T) is determined to be between the digital values (V(n-1), V(n)) corresponding to the two pixels. When the value shown is reached, the calculated position (n) of the above pixel,
The digital value (V(n-
1), V(n)) and the above threshold (T),
t=n-{V(n)-T}/{V(n)-V(n-1)
}
An image position detection method characterized in that an image position (t) indicating the threshold value (T) between the two pixels is output by performing linear interpolation calculation processing.