JPH07239930A - Image reader - Google Patents

Abstract

PURPOSE:To read an image subjected to mask processing very easily in real time through the changeover of a bias voltage. CONSTITUTION:First and second semiconductor layers 2a, 2b of a 2nd conduction type are formed on the surface of a 1st conduction type semiconductor region 1 at a prescribed interval and 1st and 2nd electrodes 3a, 3b are formed on the semiconductor layers respectively to form a pn-injection at a prescribed interval between the 1st conduction type semiconductor region and the 1st semiconductor layer and between the 1st conduction type semiconductor region and the 2nd semiconductor layer respectively. The photo detection means is formed by using a photo detection section detecting an optical current through a bias voltage applied between the 1st and 2nd electrodes as one pixel and lots of the photo detection sections to detect the light of the plural pixels simultaneously and a bias voltage is applied to each photo detection section depending on a weight function of picture processing items to be executed and the photo detection means detects a characteristic of a prescribed pixel in real time based on the output of the plural photo detection sections.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reading device,
In particular, the present invention relates to an image reading apparatus having an image processing function of performing real-time feature extraction of pixels from read outputs of a plurality of areas and outputting the extracted features as image information of the pixels.

[0002]

2. Description of the Related Art Conventionally, when an image is read using a photodetector such as a photodiode array to extract a feature, it is easy to process the information after temporarily storing the output of the photodetector such as a photodiode array in a memory. Pre-processing for feature extraction into shapes is performed, and the processed data is subjected to a recognition algorithm to obtain the required result.

That is, as shown in FIG. 17, an example of image recognition using this conventional image reading method is used to read an object using a photodetecting means 11 such as a CCD camera or a photodiode array, and the read signal is once read. No. 1 image memory 12, read a read signal of a predetermined area from the image memory, perform processing using a Robinson operator or the like (image processing unit 13), and output the image processing unit from the second image memory 14 And the image recognition 15 is performed from the stored signal.

The operation of the Robinson operator, which is one of the feature extraction methods by this image processing, will be described.

This is done by using a line sensor such as an image sensor or a photodiode array, and m in the time axis-coordinate axis plane.
The × n array of data is stocked in the first image memory 12 and sequentially read out to execute the Robinson operating of 3 × 3 data. As shown in FIGS. 18 (a) to 18 (c), the density gradient and its direction are shown. Is determined for each coordinate.

FIG. 18A shows two-dimensional data around the pixel of interest in the first image memory 2. As shown in FIG. 18B, the pixels around this pixel of interest are abcdefgh, and FIG. Perform weighting calculation as shown in c),
The largest vector is adopted as the density gradient data of the pixel of interest. In this way, the density gradient and its direction are determined, and the edge is extracted.

[0007]

However, in this method, preprocessing such as storage in the memory and reading is required before the output of the sensor is applied to the recognition algorithm, so that the system becomes complicated and time is required for processing. In addition to this, there is a problem that a large memory area is required and the device becomes large.

The present invention has been made in view of the above circumstances, and an object of the present invention is to provide an image reading apparatus which is small in size and capable of performing reading and image processing in real time.

[0009]

The feature of the present invention resides in that a second conductivity type semiconductor region is provided on a surface of the second conductivity type semiconductor region at a predetermined interval.
Of the first conductivity type by forming the first and second semiconductor layers of conductivity type and the first and second electrodes on the surfaces of the first and second semiconductor layers, respectively. Two pn junctions are formed between the semiconductor region and the first and second semiconductor layers with the predetermined distance therebetween, and a bias voltage is applied between the first and second electrodes to generate a photocurrent. The photodetector for detecting
A large number of pixels are arranged, and a light detection means configured to detect light of a plurality of pixels at the same time and a weighting function required depending on an image processing item to be executed are determined, and each pixel is weighted. A determining unit that determines a bias voltage according to a function, the bias voltage determined by the determining unit is applied to each photodetecting unit, and feature extraction of a predetermined pixel is performed based on outputs of the plurality of photodetecting units. It is provided with a drive processing unit that performs in real time and outputs this as image information of the pixel.

[0010]

According to the above construction, by changing the bias voltage, it becomes possible to easily obtain, as the light output, the light detection section weighting the actual light information. Therefore, the characteristics of a predetermined pixel are extracted in real time by adding the weighted weights to the actual light information (light intensity) in a plurality of light detectors and extracting the weighted light information as the image information of the pixel. Therefore, pre-processing is not required and no memory is required, and feature extraction can be performed very easily and in real time. In the case of the subtraction, the bias voltage may be set to −, and the weighting operation can be extremely easily performed only by adjusting the bias voltage, and the reading through the desired filter can be freely performed.

The principle of the present invention will be described. 1 and 2
As shown in (a) and (b), for example, p is formed on the surface of the n-type silicon substrate 1 so as to form a comb shape with a space W therebetween.
First and second semiconductor layers 2a and 2b made of a type diffusion layer
And the first and second electrodes 3 each formed of an aluminum layer or the like on a part of this surface.
a and 3b are formed to form one pixel. In this device, when a bias voltage is applied, a depletion layer 4 is generated depending on this bias voltage as shown in FIG. 2 (a), and an electron-hole pair is generated in the depletion layer by light irradiation. , This is drifted by the electric field and a photocurrent flows. Also, as shown in Fig. 2 (b), if the polarity of the bias is reversed, the flow of photocurrent will also be reversed. 3 and 4 show the results of measuring the relationship between the photocurrent, the light intensity, and the bias voltage. As is clear from this result, the photocurrent is proportional to the bias voltage and also to the light intensity.

The present invention has been made by paying attention to this point. Utilizing the fact that it is possible to weight the photocurrent with the bias voltage even for the light of the same intensity, and to utilize the Sobel filter and the Laplacian. A read output equivalent to that obtained by applying a filter can be obtained in real time, and a desired image can be easily read. Here, the pn junction is formed so as to form a comb shape in order to increase the ratio of the junction region per occupied area as much as possible and to increase the sensitivity.

[0013]

Embodiments of the present invention will now be described in detail with reference to the drawings.

FIG. 5 is a block diagram of an image reading apparatus capable of performing edge detection according to the first embodiment of the present invention, and FIG. 6 is a two-dimensional plan view of the light detecting means of the apparatus.

As shown in FIG. 1 and FIG. 2, this device has two pn junctions which are spaced apart from each other on a surface of an n-type silicon substrate 1 by a predetermined distance, and a bias voltage is applied between the pn junctions. A photodetection unit that applies and detects a photocurrent is defined as one pixel, and a large number of the one pixels are arranged. At the same time, a photodetection unit 21 configured to perform photodetection of a plurality of pixels and a weighting function are determined. Determining means 22 for determining a bias voltage according to a weighting function for the photodetector corresponding to each pixel.
And applying the bias voltage determined by the determining means 22 to each photodetector, and performing feature extraction of a predetermined pixel in real time based on the outputs of the plurality of photodetectors. Based on the image information extracted from the memory 24 for storing the output of the light detecting means 21, the drive processing means 23 for outputting the image information in the state where the Sobel operator mask is applied, and the output in the X direction. It is characterized by comprising a calculation processing means 25 for taking the square root of the sum of the square of the absolute value of and the square of the absolute value of the output in the Y direction and outputting this as edge strength information.

In this case, the light detecting means 21 has two pns.
Variable sensitivity light receiving element PD (1,1) composed of a junction
An array is formed by arranging 5 × 5 PD (5,5) 10, and FIG. 7 is an enlarged view of a main part thereof. Each element is shown in Figure 1.
Similarly to the above, the first and second semiconductor layers 2a, 2 made of p-type diffusion layers are formed on the surface of the n-type silicon substrate 1.
b is formed, and the first and second electrodes 3a and 3 are further formed on the surfaces of the first and second semiconductor layers 2a and 2b.
b is formed. And these switches _{SW Y1 ~SW Y5, SW X1 ~SW} X5 via the output terminal T _O, ground terminal T _G, 1V respectively, -1 V, 2
V, and it is configured so as to be switched to the bias terminal T _B of -2 V.

Next, an image reading operation using this image reading apparatus will be described.

First, the switches SW _X1 and SW _Y1 are shown in FIG.
By setting as shown in, the pixel PD (1,1)
The image of is captured.

Then, by setting the switches SW _X2 and SW _Y1 as shown in FIG. 9, the pixel to be taken in is advanced to the pixel PD (2,1).

In this way, the pixels PD (1,1) to
The normal image of the pixel PD (5, 5) is captured.

Next, a method of taking in an image on which a Sobel operator mask of 3 × 3 in the Y direction is captured will be described. As shown in FIG. 10, by connecting the switch SW _Y1 to the bias voltage terminals of 1V and 2V, the pixel P
A weight bias of 1V is applied to D (1,1) and the pixel PD (3,1), and a weight bias of 2V is applied to the pixel PD (2,1). Also, switch SW _Y3 is -1V,
By connecting to the bias terminal of -2V, the pixel PD
A weight bias of -1V is applied to (1,3) and pixel PD (3,3), and a weight bias of -2V is applied to pixel PD (2,3). Each output current at this time is SW _X1 ,
By connecting SW _X2 and SW _X3 to the output terminal T _O , addition can be performed to obtain a read output with a 3 × 3 Sobel operator mask.

Further, as shown in FIGS. 11 and 12, switches SW _Y1 and SW _Y3 are changed over to switch SW _X1 , SW _X2 and S.
3 × 3 by switching W _X3 , SW _X4 , SW _X5
Can be scanned.

Also, in the case of capturing an image on which a Sobel operator mask of 3 × 3 in the X direction is captured, it can be easily formed by switching the switches as shown in FIGS. 13 to 15.

The subject is read in this way, and the square root of the sum of the square of the absolute value of the output in the X direction and the square of the absolute value of the output in the Y direction is calculated from the read output of 25 pixels from this signal, This is output as edge strength information.

According to such an apparatus, reading and feature extraction are performed in real time, and image recognition can be performed very easily in a short time.

Further, in image processing, it is necessary to temporarily store in the memory from the read output of 25 pixels,
Only during the calculation of the square root of the sum of the square of the absolute value of the output in the X direction and the square of the absolute value of the output in the Y direction, the memory area can be small, and the device can be downsized. Is possible. This device is effective for collating still images such as seals.

When the density gradient and the density direction are obtained by the Robinson operator mask, it is not necessary to temporarily store them in the memory, and the output of the light detecting means 21 in FIG. 5 is the output of the weighting calculation as it is. By determining the bias corresponding to the mask by the determining unit 22 and applying it by the drive processing unit 23, the weighting calculation can be automatically performed.

Although the Sobel filter mask has been described in the above embodiment, the processing can be extremely easily performed by switching the bias voltage even when the differential image is captured by applying the Laplacian filter mask.
As an example of this, FIG. 16 shows an example of terminal connection when a 3 × 3 Laplacian filter mask is applied. For mask scanning, switch SW _Y1 , SW _Y2 , SW _Y3
This can be easily executed by switching _{among X1} , SW _X2 , SW _X3 , SW _X4 , and SW _X5 .

[0029]

As described above, according to the present invention, it is possible to take in a masked image very easily and in real time by switching the bias voltage.

[Brief description of drawings]

FIG. 1 is an explanatory view of the principle of an image reading apparatus of the present invention.

FIG. 2 is an explanatory view of the principle of the image reading apparatus of the present invention.

FIG. 3 is a diagram showing a relationship between photocurrent and irradiation light intensity in the device.

FIG. 4 is a diagram showing a relationship between a photocurrent and a bias voltage in the device.

FIG. 5 is a block diagram showing an image reading apparatus according to an embodiment of the present invention.

FIG. 6 is an explanatory diagram showing an image reading apparatus according to an embodiment of the present invention.

FIG. 7 is an enlarged view of a main part of the device.

FIG. 8 is a diagram showing a normal image capturing state using the apparatus.

FIG. 9 is a diagram showing a normal image capturing state using the apparatus.

FIG. 10 is a diagram showing a captured state of an image when a 3 × 3 Sobel operator mask in the Y direction is applied using the same device.

FIG. 11 is a diagram showing a captured state of an image when a 3 × 3 Sobel operator mask in the Y direction is applied using the same apparatus.

FIG. 12 is a diagram showing a captured state of an image when a 3 × 3 Sobel operator mask in the Y direction is applied using the same apparatus.

FIG. 13 is a diagram showing a captured state of an image when a 3 × 3 Sobel operator mask in the X direction is applied using the apparatus.

FIG. 14 is a diagram showing a captured state of an image when a 3 × 3 Sobel operator mask in the X direction is applied using the apparatus.

FIG. 15 is a diagram showing a captured state of an image when a 3 × 3 Sobel operator mask in the X direction is applied using the apparatus.

FIG. 16 is a diagram showing a state of capturing an image when a Laplacian operator mask is applied to the apparatus.

FIG. 17 is a diagram showing a conventional image recognition device.

FIG. 18 is an explanatory diagram of image recognition using a Robinson operator of a conventional example.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 n-type silicon substrate 2a, 2b semiconductor layer (p-type diffusion layer) 3a, 3b electrode 11 light detecting means 12 first image memory 13 image processing section 14 second image memory 15 image recognition 21 light detecting means 22 determining means 23 Drive Processing Means 24 Memory 25 Arithmetic Processing Means

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Office reference number FI technical display location H04N 1/40 101 D

Claims (1)

[Claims] 1. A first conductivity type semiconductor region and a second conductivity type first and second semiconductor layers are formed on a surface of a first conductivity type semiconductor region at predetermined intervals, and the first and second semiconductors are formed. By forming a first electrode and a second electrode on the surface of the layer, respectively, the predetermined distance is provided between the semiconductor region of the first conductivity type and the first and second semiconductor layers.
One pn junction is formed, a bias voltage is applied between the first and second electrodes, and a photodetection section for detecting a photocurrent is defined as one pixel. Light detection means configured to perform light detection, and determination means for determining a weighting function required according to an image processing item to be executed and determining a bias voltage corresponding to the weighting function for each pixel. And applying the bias voltage determined by the determining means to each photodetector, and performing feature extraction of a predetermined pixel in real time based on the outputs of the plurality of photodetectors. An image reading apparatus comprising: a drive processing unit that outputs information.