JP2008131609A - High sensitive and high-speed electronic determining device for competition using charge multiplying device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a high sensitive, high-speed, and improved electronic determining device for competitions using a charge multiplying device, and to provide a system. <P>SOLUTION: The electronic determining device for competitions uses a line sensor having a charge multiplying section behind a CCD element for increasing sensitivity and restraining noise, thus enabling images to be captured even in a low illuminance condition in the evening or at night and/or a high-speed photographing condition of 10,000 lines/second, where the images cannot be captured by a normal CCD or CMOS sensor. The electronic determining device for competitions is used to connect a CCD electronic camera control unit of the electronic determining device to a personal computer for display through a G Ethernet(R). <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an improved high-sensitivity and high-speed electronic determination device in competitions (race) such as horse racing, bicycle racing, boat racing, and auto.

2. Description of the Related Art Conventionally, a system is known in which a goal sensor is continuously imaged on a goal line in horse racing and the like, and arrival order determination and time display of horse racing are performed from the captured images.
A line sensor (linear CCD sensor) is used in the electronic judging device for competition, and the line sensor images the line on the goal line of the race track at a certain slit interval, These images are continuously displayed on the monitor to obtain an arrival order determination image.

  This is a CCD line sensor that uses a slit camera to shoot a film continuously at approximately the same speed as the object (eg, horse) and display it as an arrival order determination image. And electronically realized by a monitor.

  However, the sensitivity of the conventional line sensor is low, and the image becomes blurry when taken in low-light conditions such as the recent night game or the cloudy evening race, the sharpness is lowered and the resolution is deteriorated. For example, there are many unsatisfactory cases in determining the arrival order of the difference in the nose.

In addition, a highly accurate electronic determination device suitable for high-speed processing is desired for the determination of a close arrival order such as a nose difference on a goal line such as a horse race passing at high speed.
In the conventional electronic determination device, a general CCD or CMOS element is used for each light receiving element of the line sensor that becomes the eye. Therefore, the sensitivity of the line sensor is not sufficient in a low illuminance state such as evening or night, and the luminance is low. Therefore, it is difficult to see the image for determination with high noise, which hinders determination of arrival order.

In addition, the sharpness is not always sufficient as a high-speed photographed image, and there is a situation where the reliability that replaces the conventional photo determination cannot be obtained.
On the other hand, a CCD sensor itself having a charge multiplying element is well known in the art, but there is no known one that applies this to an electronic determination system for competition.
Japanese Patent No. 3224735 JP-A-6-276526 JP 2003-347317 A USP 5,337,340

  SUMMARY OF THE INVENTION An object of the present invention is to provide an improved electronic determination apparatus and system for competition using a charge multiplying element and improved in sensitivity and speed.

In order to achieve the above object, the electronic determination device for competition according to the present invention uses a line sensor having a charge multiplying portion behind the CCD element to increase sensitivity and suppress noise, thereby reducing the noise from a normal CCD or CMOS sensor. In this case, it is possible to take an image even in a low illumination condition in the evening or night and / or in a high-speed photographing condition of 10,000 lines / second.
Further, the line sensor is a line sensor having an overflow drain, and the occurrence of smear is reduced.

In addition, a Peltier device is mounted on the back surface of the line sensor and is cooled to below 0 ° C., thereby reducing the occurrence of thermal noise and improving the image quality of the determination image.
Further, the output image of the line sensor having the charge multiplying function is smoothed only in the position direction (longitudinal direction) by the weighted moving average or the median value selection method, and the smoothing process is completely performed in the time axis direction. Impulse noise that is randomly generated in the charge multiplication section without being performed is removed.

  Alternatively, the electronic determination system for competition according to the present invention uses a line sensor having a charge multiplying portion behind the CCD element to increase sensitivity and suppress noise, so that an evening image that cannot be captured by a normal CCD or CMOS sensor. Using a competition electronic judging device capable of imaging even under low light conditions at night and / or at a high speed photographing condition of 10,000 lines / second, between the CCD electronic camera control device of the electronic judging device and a display personal computer Are connected by G Ethernet.

  Furthermore, a frame buffer unit is provided inside the electronic camera control device, and the image of the line sensor is constructed and transmitted in a format such as a standard two-dimensional image in the frame buffer unit, so that it is supported by G Ethernet as a standard. The video information and digital information (elapsed time from start, date, venue, etc.) of the line sensor are transmitted within the normal rules of G Ethernet based on the rules of 2D camera transmission, and the display personal computer The image information and digital information of the line sensor transmitted inside are reconstructed and displayed.

  In the apparatus of the present invention, by using a line sensor having a charge multiplying portion behind the CCD element as the line sensor, sensitivity was improved, noise was suppressed, and a good electronic determination image was obtained even at low luminance.

  Similarly, in the apparatus of the present invention, a line sensor having a charge multiplying portion behind the CCD element is used as a line sensor, thereby increasing sensitivity and suppressing noise, thereby achieving high speed (10,000 lines / second). ) But a good electronic determination image was obtained.

  Therefore, according to the improved electronic device and system for competition with high sensitivity and high speed according to the present invention, low light conditions in the evening and night and / or 10,000 which cannot be imaged by a normal CCD or CMOS sensor. Even under high-speed shooting conditions of line / second, it was possible to take images by increasing sensitivity and suppressing noise.

  This increases the convenience of the electronic determination device and improves the reliability of the electronic determination device by improving the accuracy of the determination.

  In the electronic determination device of the present invention, a line sensor having a charge multiplying portion behind the CCD element is used as the line sensor. By using the charge multiplication type line sensor, sensitivity can be improved, noise can be suppressed, and a good electronic determination image can be obtained even at low luminance.

  On the other hand, in the conventional electronic determination device, a general CCD or CMOS element is used for each light receiving element of the line sensor that becomes the eye, so that the amount of light is insufficient in high-speed shooting, and about 2,000 lines / second is obtained. Although it was a practical limit, in the electronic determination device of the present invention, by using a line sensor having a charge multiplying portion behind the CCD element as a line sensor, sensitivity is increased and noise is suppressed, so that high speed (10 , 000 lines / second), an excellent electronic determination image can be obtained.

  The present invention is a charge sensor developed by Texas Instruments, for example, disclosed in Japanese Patent Application Laid-Open No. 2003-347317 and US Pat. It is characterized by using a type CCD sensor (hereinafter referred to as an IMPACTRON device).

  The IMPACTRON device is a CCD with a new structure having a boost structure that induces secondary electrons to charge during transfer in a chip. The device amplifies the charge before the read amplifier and theoretically realizes an ultra-sensitive CCD video camera that detects even one particle of light.

Details of the operation principle are described in Japanese Patent Application Laid-Open No. 2003-347317 and US Pat. No. 5,337,340, and the outline thereof is as follows.
FIG. 1 is a schematic diagram of the conventional CCD sensor, and FIG. 2 is a schematic diagram of an electron multiplying sensor used in the present invention.

  A conventional CCD sensor has a readout amplifier 2 connected to a light receiving unit (CCD) 1 as shown in FIG. 1, but an electron multiplication type sensor is located at the rear stage of the light receiving unit 3 as shown in FIG. In addition, a charge multiplier 4 is provided in front of the conventional read amplifier 5.

  As shown in the conceptual diagram of FIG. 3, the electron multiplying type sensor 3 is charged by the charge multiplying unit 13 when the charges received and photoelectrically converted by the sensor unit 10 are accumulated and transferred (11 and 12 in FIG. 3). After being multiplied, it is input to the read amplifier. FIG. 4 shows the detailed structure. In the schematic diagram 14 of FIG. 4, the signal charge 15 of P2 is transferred to the deep potential 16 formed of P4. At the time of transfer to the P4 gate, an impact ionization phenomenon (charge collides with Si and generates electrons and holes) due to a potential difference is used to take out secondary electrons. Further, by repeating this in this impato ionization chip, it is possible to multiply the signal charge without superimposing readout noise.

  Therefore, as shown in FIG. 5, the electron multiplication type line sensor inputs the signal after being multiplied to the CCD readout amplifier, so that the amplification is completed before voltage conversion of the electric charge. The influence of internal noise is very small.

Hereinafter, characteristics of the IMPACTRON device are listed as follows.
A. Built-in charge multiplier Built-in on-chip charge multiplier that directly amplifies the charge stored on the CCD.
B. Unlike a photoelectron acceleration type multiplication mechanism that requires several KV, a high voltage is not required. Amplification of about 100 times or more is possible with a Boost voltage amplitude of 15V.
C. High resolution There is no fluorescence emission or fiber plate on the front of the device, and photoelectrons are received directly, so there is no afterimage at high resolution.
D. Since amplification is performed before the output amplifier, the noise becomes low.
E. Colorization is possible.
F. The built-in cooling Peltier realized a reduction in device dark output.
G. This CCD manufactured by TI can handle from ultraviolet to near infrared.
H. Because it is a solid state device, it is small and highly reliable.
I. There is no problem of burn-in, and high reliability and long life can be realized.

Hereinafter, each item will be described in detail.
・ Overflow drain In the conventional electronic determination device, when the light exceeding the limit amount is accumulated inside the light receiving element due to the strong light incident on each light receiving element of the line sensor, the response to the surplus charge Since the sensor was not used, the surplus charge flowed into the adjacent pixel, and an appropriate image could not be acquired.

  As a result, as a conventional electronic determination device, strong light is received from positions such as a horseshoe or a bicycle rim, and whitish streaks appear as an image in the line direction, which hinders determination. In the present invention, a line sensor is prepared adjacent to each light receiving element, in order to eliminate the flow to the adjacent pixels by preparing a path for releasing surplus charges, and flowing charges exceeding the limit amount to the path. By using it, even if strong light is incident, excessive charges are eliminated and a good electronic determination image can be obtained.

FIG. 7 shows the specific configuration.
Smear is a phenomenon in which when a very bright subject exists in the imaging area, the amount of charge that is photoelectrically converted during vertical transfer cannot be ignored, and whitish streaks appear vertically all around the high-luminance subject. .

  When a subject with very high brightness exists in the imaging area, the amount of charge that is photoelectrically converted during vertical transfer cannot be ignored, and whitish streaks appear vertically throughout the subject with high brightness. The photograph in FIG. 6 shows the experimental results with input light 4000 times the saturation light quantity.

  In this way, smear is a phenomenon that occurs when strong light enters the image pickup surface of the CCD, and surplus charges generated there overflow to pixels around that pixel, resulting in an image in which light oozes out to the surroundings. Become. This is suppressed by providing an anti-blooming mechanism in the CCD element and discharging excess charges. The company TI employs a LOD (lateral overflow drain) mechanism 22 shown in FIG.

・ G Ethernet Conventionally, the case where the camera part in charge of imaging of the electronic judging device and the personal computer part in charge of display are transmitted by analog video information and the case of transmission by digital transmission developed for video signals such as IEEE1394 are mainly used. It was.

  This is a method of acquiring an analog output signal from a camera with a camera input interface installed in a personal computer, and is the most popular method that has been used since the past. The method of connecting with the digital signal of this IEEE1394 standard is as follows.

  This method is a method of transmitting image information between a camera, a video device, and a personal computer first proposed by Apple Inc. in the United States, and this method is currently in practical use only up to 400 MBit / second, mainly NTSC or PAL. It was developed for the transmission of image signals of the same level as the conventional broadcast standard (cannot be applied to high-definition video that started in Japan). Therefore, a high-speed camera that far exceeds 400 MBit / sec. Is not applicable.

Reference: Amount of image information generated by the current camera 10,000 lines / second * 4000 pixels / line * 3 (RGB)
= 120 MByte / sec = about 1200 MBit / sec (1 Byte = 8 Bit should be converted, but since various header information is added before and after the packet on the transmission path, 1 Byte = about here in consideration of this) (Converted to 10Bit)

In the present invention, as shown in FIG. 8, it is proposed to transmit image transmission from a camera to a personal computer using G Ethernet.
G Ethernet uses a CSMA / CD (Carrier Sense Multiple Access / Collision Detection) method, and transmits a frame by checking whether a data path is free before transmitting the frame. On the other hand, since the video signal is high-speed and cannot be stopped halfway, the electronic determination system of the present invention includes a frame buffer unit inside the electronic camera control device, and the image of the line sensor is displayed in the frame buffer unit. It is constructed and transmitted in a format such as a standard two-dimensional image. By reconstructing and transmitting the data as if it were a standard 2D image, the video information + digital information (from the start) of the line sensor is maintained, while complying with the 2D camera transmission rules supported by the G Ethernet standard. (Elapsed time, date, place, etc.) are sent to the PC by transmitting within the normal G Ethernet rule range (eg 240 x 240 pixels), and the application software inside the PC is transmitted in the normal 2D image format. By reconstructing the image, the image of the anomalous line sensor for G Ethernet is received and displayed.

  In addition, the conventional general method is to connect multiple PCs with an Ethernet (registered trademark) standard transmission interface, and since each PC has a unique address on the Ethernet, a message is sent from the sender to the intended receiver. Most connections are connected to the Internet using this Ethernet standard. Ethernet started at a transmission speed of about 10 MBit / second in the early days, but the transmission speed increased to 100 MBit / second and 1000 MBit / second. In recent personal computers, there is a personal computer that has a function of the transmission speed of 1000 MBit / second as a standard. It has increased. This Ethernet having a speed of 1000 MBit / second is simply called G Ethernet.

  In Ethernet, the sender basically sends out the packet that the sender wants to transmit on the Ethernet with the address of the receiver, and if there is a receiver with the corresponding address on the same Ethernet, the transmission that is established when the receiver returns a response. This method is also a transmission method that allows another sender to send to another receiver at the same time (allows collision), so a terminal that wants to transmit certain data at a certain time like a camera It is not suitable as a means for transmitting image information from the point of view that the information is surely passed to the receiver. However, most of the personal computers in the world are equipped with Ethernet, which is an inexpensive and high-speed transmission means, as standard equipment, and are connected to the Internet etc., so it is very convenient to use this communication means.

Hereinafter, a specific connection method in the G Ethernet according to the present invention will be described.
Note that a color having a resolution of approximately 256 (horizontal direction) x 240 (vertical direction) or 640 (horizontal direction) x 480 (vertical direction) when a conventional two-dimensional camera (area sensor) is used. There are methods for transmitting a two-dimensional camera and a monochrome camera having a resolution of 1024 * 1024 using G Ethernet.

256 pixels * 240 lines * 60 frames / second * 3 (RGB) = about 11 Mbyte / second 640 pixels * 480 lines * 30 frames / second * 3 (RGB) = about 28 Mbyte / second 1024 pixels * 1024 lines * 30 frames * 1 (Monochrome) = approx. 32 MByte / sec This is practical because it has enough transmission capacity compared to the G Ethernet transmission capability, but in reality, application software that increases the load is executed on a personal computer. If this is the case, this is a method of knowing that the mistake has been made.

  However, the system of the present invention is an electronic determination device for competition, and even one line is not allowed to be lost. Therefore, in this system, a buffer 73 is provided in the camera control unit as shown in FIG.

(1) The analog image signal output from the line sensor in the camera head 70 is sent to the camera control unit 80, and is subjected to AD conversion by performing minimum analog processing (GAIN adjustment or the like).
(2) The AD converted line sensor image is subjected to the necessary image processing (noise removal, gamma conversion, white balance adjustment, etc.), information such as time information and venue information is added to the top of the image, and the personal computer is used as it is. Since it will be in the raster direction (horizontal direction) on the personal computer, it will be necessary to convert it to the vertical direction used for competition, and this part will be converted to vertical and horizontal (72).
(3) The CPU 81 in the camera control unit instructs each unit according to an instruction from the personal computer via the G Ethernet, or a button instruction such as time start, recording start, recording pause, recording end directly connected to the camera control unit. As shown in FIG. 12, the image information is recorded until an instruction to stop recording is received in a buffer memory 73 sufficient to store one race worth of the image that has been subjected to the vertical / horizontal conversion.
(4) When it is confirmed that the above-described image information that can be sent out by the G Ethernet 82 is stored in the buffer memory 73, the “one-dimensional to two-dimensional conversion and sequential number addition circuit” 74 receives the image information from the buffer memory. Start reading and reassemble from 1D information to 2D information. At this time, a sequential number is added to the head as shown in FIG.
(5) Image information assembled into two-dimensional information is sent to a personal computer via the G Ethernet 82.

(6) The application software on the personal computer 79 confirms the leading sequential number each time it acquires image information from the apparatus as if it were sent as 2D camera information from G Ethernet (78 in FIG. 11). The sequential number is 0 immediately after the start of recording of the corresponding race, and the value after that is incremented continuously by 1 (it is safer if a checksum is added to the sequential number). If the sequential number of the image acquired by the application software is n and the acquired number is n + 2, then n + 1 images have been lost on the transmission path. If n + m, then images from n + 1 to n + m−1 have disappeared on the transmission path. In this case, the application software sends the disappearance information to the CPU 81 of the camera control unit via G Ethernet. The received CPU 81 resends the corresponding line, but resends a part of the sequential number with a flag indicating that it is a resent image. By doing this, the image sent normally is placed at a predetermined position in the storage memory in the personal computer, and the position of the storage memory can be specified later even if the image is sent later by resending, so it is inserted at that location. .
If the value is not expected in the sequential number check (78), the application software is discarded. (For example, an image that is expected to be n + 1 but does not meet the checksum, such as nm)

(7) The reason why the above-described (6) is performed is best performed in accordance with the relationship between the high-speed image transmission speed and the response speed of the application software. Assuming that a two-dimensional image is a color image interpretation of 256 × 240 pixels as a two-dimensional image, if the image is sent with a clock of about 30 MHz, one image reaches the personal computer in about 6 mS. On the other hand, even if the response of the application software is fast, the execution right can be received from the OS only once every several milliseconds. Once the execution right is received, high-speed work of less than mS is possible.

  The personal computer 79 also secures a sufficiently large buffer for temporarily storing images received from the Ethernet, and stores the images sent one after another from this apparatus by leaving it to the driver software under the OS. The application software sometimes checks if this buffer has accumulated more than half, and if it exceeds that, instructs the camera control unit to stop transmission via G Ethernet. Or, when it is almost gone, it instructs the camera control unit to resume transmission. (In general, it should be indicated by detecting less than 5% of the buffer size)

The camera control unit continuously sends out image frames until there is a transmission stop command from the application software. The driver software sequentially stores in the buffer in the personal computer 79 in response to this transmission, but a certain image frame is lost depending on the execution load of the personal computer (G Ethernet does not guarantee). Therefore, it is necessary to confirm whether all the images have arrived correctly by adding the sequential numbers as described above. Thus, although it is inexpensive and fast from a device that transmits at a constant speed such as a camera, it can be safely transmitted even by G Ethernet where image loss is performed.
As described above, the video signal is continuously and stably transmitted despite using G Ethernet generally used in the world.

FIG. 10 shows a specific example (hereinafter referred to as a CDV-X system) of the competition electronic determination system of the present invention.
In FIG. 10, the CDV-X system is centered on a memory box (hereinafter referred to as DCU) 56, sensor camera 50, dedicated CCD line 59, optical fiber 57, Gigabit Ethernet (hereinafter referred to as G Ethernet) 58, PC (personal computer). ) 55, a main / sub console 60.61, and a liquid crystal monitor 62 having a UXGA resolution (1600 × 1200).

  The DCU 56 conditions an analog signal from the camera head and performs A / D conversion. The image data is stored in a buffer memory having a storage capacity of 2 GB in the DCU, the format is converted and transferred to the PC 55. The PC 55 displays the received race data on the PC screen, performs enlargement / reduction and image processing, and finally records the data as digital data on the hard disk.

Since there is one buffer memory in the DCU 56, even if an accident occurs on the PC 55 side, it is possible to reacquire race data by going back to the previous race.
Control communication between the PC 55 and the DCU 56 is performed through a serial communication line in the camera link interface. The baud rate is 38.4 Kbps.
An analog process board and a digital processing board are provided in the DCU 56, and the role of each board is as follows.

・ Analog process board Generates clocks and synchronization signals necessary for camera operation.
The CPU (SH-4) in the DCU 56 receives the setting from the PC 55 and generates a timing signal based on the setting.

An analog signal connected from the camera head is input to a 10-bit A / D converter through processing such as gain adjustment, offset adjustment, and filter, and is converted into digital data.
The data is input into the Altera FPGA and is connected as a LVDS signal to the digital processing board through various signal processing.

-Digital processing board The digitized image data is received from the analog process board, and first, vertical / horizontal conversion is performed to change the vertical scan data to raster scan data. However, conversion is performed in blocks for every 16 lines of vertical scan due to buffer memory capacity and convenience when sending out the camera link.

  That is, the data for 16 lines is converted into raster scan data of 16 horizontal pixels × 3600 vertical pixels. At this time, a time code is embedded in the image for each vertical line.

Next, the image data is written into a buffer memory having 2 GB.
Further, this data continues to be held except when the next race data is acquired and when the power is turned off. Therefore, even if the transfer to the PC fails for some reason, the maximum transmission to the PC is possible.

The data written in the buffer memory is sequentially read out and sent to the camera link adapter as 256 × 226 pixel area scan camera data.
In addition, this board is equipped with an SH-4 CPU manufactured by Renesas and controls external input and trigger signals from the PC to perform various control of image capture, internal settings in each place, communication management with the camera head CPU, and the like. .

・ Camera Link ← → Gigabit Ethernet Media Converter The race data output from the digital processing board via the camera link interface is input to the “Camera Link to Gigabit Ethernet” converter 52 manufactured by Preora and output as a Gigabit Ethernet signal.

The maximum transfer rate is 100 MByte / Sec, and this device uses a maximum speed of about 60 MByte / Sec.
Gigabit Ethernet ← → Optical fiber media converter The Gigabit Ethernet signal is further input to the Gigabit Ethernet to optical fiber converter 53 and output as an optical fiber signal.
The signal is single mode transmission and the reach is 1 km or more.

-Element cooling In the conventional electronic determination device, each light receiving element of the line sensor that is the eye is a general CCD or CMOS element, and the element is not cooled. Noise was the dominant image for determination, which hindered determination.

  In the apparatus of the present invention, as shown in FIG. 9, a Peltier element 44 is mounted on the back surface of the element 41 of the line sensor and cooled to below about 0 ° C., thereby reducing the temperature of the element and reducing the generation of thermal noise. The determination work is simplified by improving the image quality of the determination image.

・ Quantum noise countermeasures The output image from a line sensor with a charge multiplication function cannot be ignored as the multiplication ratio is increased. There were many images, which hindered judgment.

  In the apparatus of the present invention, the output image from the line sensor having the charge multiplying function is a value indicating a median value only in the column direction so as not to affect the time direction and compared with a pixel near a certain target pixel. By substituting for it, the process of removing noise was performed and the image quality was improved to make the judgment work easier.

-Measures against light shot noise When the amount of photons in a dark subject decreases, the level of light fluctuations is captured, and a CCD camera with a fast response causes image quality degradation due to random noise. This is because the characteristic of the CCD having no afterimage has an adverse effect.

  In order to eliminate the adverse effects caused by the speed of response, the image quality can be improved by connecting a smooth process such as signal averaging to the subsequent stage, similar to the characteristics of the phosphor screen of the image intensifier.

This specific aspect is as follows.
First, what is important in the present invention is that the processing in the time direction is not performed in the smoothing.

  As a countermeasure against shot noise of a normal sensor, the upper and lower, left and right pixels in a certain area are taken and the average value is adopted. As a result, the continuity of the pixels becomes smooth and clickable noise is removed.

However, since the main purpose of the line sensor of the present invention is to determine the arrival order on the goal line, taking the average value before and after on the time axis is not preferable because the accurate position of the image is lost.
Therefore, in the present invention, the accuracy of determination is prioritized over the image quality, and the smoothing process is performed only between the upper and lower positions (sensor longitudinal direction), and the pixels shifted on the time axis, that is, consecutive pixels with different scans. No smoothing in between. This improves the determination accuracy on the time axis.

This smoothing process digitizes the image from the sensor with D1, D2,... Dn,.
For example,
I. Weighted moving average (example is the same for 3 position method, 5, 7, 9 position)
DnNEW = (a * Dn-1 + b * Dn + c * Dn + 1) / (a + b + c)
Dn: There are three types, R, G and B. Input image data indicating a value of 0 to 4096 (when a 12-bit AD converter is used)
a, b, c: Any value in the range of 0 to 256 (mainly used to remove thermal noise)
B. Simple median selection method (example: 3-position method, the same procedure for positions 5, 7, 9)
GnNEW = SORT (Gn-1, Gn, Gn + 1)
RnNEW = SORT (Rn-1, Rn, Rn + 1)
BnNEW = SORT (Bn-1, Bn, Bn + 1)
The SORT function here is a function that rearranges the given DATA in the order of large or small and uses the median as the value on the left side. For example, assuming that Gn-1 = 100, Gn = 120, Gn + 1 = 110, and arranging them in ascending order, 100 (Gn-1), 110 (Gn + 1), 120 (Gn) are used, and 110 (Gn + 1) is adopted as the median value. The In the above formula, the description is based on a color image.
(Used to remove random impulse noise generated at the charge multiplier.)

C. Green standard median selection method GnNEW = SORT (Gn-1, Gn, Gn + 1)
RnNEW = Rn at a position that is a candidate for GnNEW
BnNEW = Bn at a position that is a candidate for GnNEW
The above formula is described on the assumption of a color image.
(Used to remove random impulse noise generated at the charge multiplier.)
Etc. are used.

  According to the present invention, it is possible to perform competitions at night or in the evening in low light conditions, high-speed shooting, and high resolution, so that the reliability of the electronic determination device, which has been low in reliability, is improved, and the conventional photo The possibility that the electronic determination device of the present invention is used for the arrival order determination on the goal line that relied on the determination has increased.

  In addition, the arrival order determination work becomes accurate and easy and is streamlined, and there is an effect that the photo determination work that relies on manual work is reduced. Furthermore, the computerization and recording operations are greatly simplified, and the industrial applicability of the above-described electronic determination device of the present invention with high resolution capable of high-speed imaging at low illumination conditions such as nighttime. Is expensive.

It is a schematic diagram of a conventional CCD sensor. It is a schematic diagram of an electron multiplication type sensor used in the present invention. It is a conceptual diagram of an electron multiplication type sensor. It is a figure which shows the structure of the detail of an electron multiplication type sensor. It is explanatory drawing of the operation principle of an electron multiplication type sensor. It is a figure which shows the experimental result in the input light 4000 times the saturation light quantity. It is explanatory drawing of an overflow drain. It is a connection diagram of G Ethernet. It is a figure which shows the cooling structure of an electron multiplication type sensor. It is a figure which shows the specific Example of the electronic determination system for competitions of this invention. It is a conceptual diagram of G Ethernet of the electronic determination system for competitions of the present invention. It is explanatory drawing of the aspect conversion of the electronic determination system for competitions of this invention. It is explanatory drawing of the receiving side personal computer process of the electronic determination system for competitions of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Light-receiving part 2 Amplifier 3 Light-receiving part 4 Charge multiplication part 5 Amplifier 10 Sensor part 11,12 Charge transfer part 13 Charge multiplication part 20,21 Light-receiving element 22 Overflow drain 30,35 Camera head 31,36 Camera control part 32, 37 PC 33, 38 Monitor 41 Light receiver 44 Peltier element 50 CCD camera 55 PC
56 DCU
70 Camera head 73 Buffer memory 79 PC
80 Camera control unit 81 PC

Claims (6)

  By using a line sensor having a charge multiplying part behind the CCD element to increase sensitivity and suppress noise, low illumination conditions in the evening and night that cannot be imaged by a normal CCD or CMOS sensor, and / or 10,000 An electronic determination device for competition, characterized in that imaging is possible even under high-speed shooting conditions of line / second.   2. The electronic determination device for competition according to claim 1, wherein the line sensor is a line sensor having an overflow drain, and generation of smear is reduced.   The competition according to claim 1, wherein a Peltier device is attached to the back surface of the line sensor and cooled to below 0 ° C to reduce the generation of thermal noise and improve the image quality of the judgment image. Electronic determination device.   The output image of the line sensor having the charge multiplication function is smoothed only in the position direction (longitudinal direction) by the weighted moving average or the median value selection method, and the smoothing process is not performed at all in the time axis direction. 2. The game electronic determination device according to claim 1, wherein impulsive noise generated at random in the charge multiplication section is removed.   By using a line sensor with a charge multiplier behind the CCD element to increase sensitivity and suppress noise, low-light conditions in the evening and night that cannot be imaged with a normal CCD or CMOS sensor, and / or 10, A competition electronic judging device capable of imaging even under a high-speed shooting condition of 000 lines / second, and a CCD electronic camera control device of the electronic judging device and a display personal computer are connected by G Ethernet. Electronic judgment system.   A frame buffer unit is provided inside the electronic camera control device, and the image of the line sensor is constructed and transmitted in a format such as a standard two-dimensional image in the frame buffer unit, and is supported by G Ethernet as a standard. Based on the rules of 2D camera transmission, video information and digital information (elapsed time from start, date, location, etc.) of the line sensor are transmitted within the normal rules of G Ethernet, and inside the display personal computer. 6. The game electronic determination system according to claim 5, wherein the transmitted video information and digital information of the line sensor are reconstructed and displayed.