JP2006324614A - High dynamic range, pixel-periphery recording imaging device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an imaging element, having a region for temporarily storing signal charges at a certain threshold or higher, by using a useless space necessarily produced in design in a pixel-periphery recording imaging device for ultra high-speed cinematography. <P>SOLUTION: Only image signals at a threshold or lower are amplified by collision ionization multiplication to be read, then an image signal at the threshold or higher is read without multiplication. Accordingly, ultra-high sensitivity, ultrahigh-speed cinematography having a large dynamic range is achieved. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

Detailed Description of the Invention

  High-speed phenomenon shooting and slow motion playback.

The inventors of the present invention have developed a pixel peripheral recording type image sensor as an image sensor for an ultra-high speed video camera. This is a technology capable of creating 100 or more image signal recording areas that are compact and low-noise in each pixel. With this technology, by simultaneously recording image signals in parallel on all pixels during shooting, 1 We have developed a video camera that can shoot at over 1 million frames per second.
JP-A-2001-345441 (P2001-345441A), high-speed imaging device and high-speed imaging device, published on December 14, 2001

  The concept of this image sensor is shown in FIG. That is, electrons generated in the photodiode 101 are used as an image signal, and the image signal is transferred to the CCD 102 linearly extending downward therefrom. This CCD is used as an image signal recording area. Since all the photodiodes have such an image signal recording area, it is possible to record image signals in parallel for all the pixels.

  A drain 103 is provided at the end of each image signal recording area. During imaging, a new image signal is continuously sent from the photodiode to the recording area, and an old image signal is continuously discharged from the drain to the outside of the image sensor. , The latest continuous image signal is always stored in the image sensor.

  In FIG. 1, the number of CCD memory elements in each image signal recording area 102 is 15, and continuous super-high-speed continuous overshooting of 15 images is possible. The actual number of CCD memory elements in the device is 100 or more, which enables moving image observation for 10 seconds or more at a reproduction speed of 10 sheets / second.

  After the photographing is finished, the image signal stored in the image sensor is sequentially transferred to the read horizontal CCD 105 formed outside the light receiving surface through the read vertical CCD 104, and further read out to the outside of the image sensor through the read amplifier 106 at the end. . The read signal is converted into a digital signal by the AD converter 111 outside the image sensor. In FIG. 1, in addition to the elements described in the above description, some elements are added, which will be described later.

  As shown in FIG. 1, the readout vertical CCD 104 is connected to the lowermost section of the CCD memory 102 of each pixel.

  In order to realize such a structure, as shown in FIG. 1, a large useless space 108 is inevitably generated by design between the end section 107 of the readout vertical CCD and the readout horizontal CCD 105.

On the other hand, Heinechek invented a collision ionization multiplication type CCD, which is an efficient image signal multiplication device unique to a CCD type imaging device. This is called CCM (Charge Carrier Multiplier). Although it was originally called CMD, it is now called CCM because there was another patent named CMD. The principle is briefly described as follows.
U. S. Patent No. 5337340 Hynecek, Charge multiplexing detector (CMD) suite for small pixel CCD image sensors

  When the electric field for transferring charges on the CCD is increased beyond a certain limit, the electrons are accelerated at high speed, and they collide with surrounding atoms to generate secondary electrons. Such a phenomenon is called impact ionization. Since the number of secondary electrons generated is a random amount and becomes noise, the CCD is usually designed under such a condition that such a phenomenon does not occur. On the other hand, if very weak ionization multiplication is repeated many times on a CCD having several hundred stages, the average number of generated electrons increases in proportion to the number of collisions, that is, the number of stages of CCM, and the standard deviation is the square root of the number of collisions. It increases in proportion to Since the signal S is an average value and the noise N is proportional to the standard deviation, the S / N ratio, which is the ratio, increases in proportion to the square root of the number of stages (S / N ratio = signal intensity / noise intensity = average number of generated electrons / The standard deviation = the number of CCM stages / the square root of the number of CCM stages = the square root of the number of CCM stages).

  FIG. 1 shows a CCD 109 added to a normal CCD image sensor. Thus, the CCM is inserted between the read horizontal CCD 105 and the read amplifier 106.

  When shooting a very dark subject such as astronomical photography, the image signal becomes very small. At this time, the signal can be amplified by the CCM to be higher than the noise level of the read amplifier. In this way, CCM is a very effective means for shooting under conditions where the amount of light incident on each image is small for shooting under weak light or high-speed shooting.

  However, not only small signals but also large signals are amplified simultaneously by the CCM. When the amplified signal overflows beyond the capacity of the readout amplifier, it cannot be measured.

  Therefore, there is a method to secure a large dynamic range as well as high sensitivity by providing a drain in the readout horizontal CCD and transferring it to the readout amplifier via a separate transfer path without amplifying the overflowing charge. It has been devised.

    Usually, such an improvement complicates the structure around the horizontal CCD and CCM. For this purpose, a relatively large space is required.

Problems to be solved by the invention

  Broadly speaking, there are the following two problems.

  First, in the pixel peripheral recording type imaging device, as shown in FIG. 1, a large useless space 108 is generated between the end section 107 of the readout vertical CCD and the readout horizontal CCD 105.

  Secondly, in the CCM, it is necessary to amplify and output only a small image signal, and to add a function of reading the large image signal without reducing the CCM amplification rate or CCM amplification. This achieves high sensitivity and a large dynamic range at the same time.

  In order to solve the second problem, it is necessary to solve some specific problems. For example, if an area for temporarily storing a large signal charge is provided on the side surface of the CCD, after the small signal charge is read, the temporarily stored large signal charge is returned to the original CCD transfer path. Need to be read from. For this purpose, it is necessary to add a function of temporarily emptying the transfer path in order to return the large signal charge to the CCD transfer path after reading out the small signal charge.

  These must be realized with the simplest possible structure. In particular, it is necessary to make the simplification as much as possible for a structure on an image sensor in which space is limited and it is necessary to avoid making extra elements in order to minimize noise contamination. In particular, increasing the number of metal wires increases the possibility of short-circuiting and may lead to a significant decrease in yield rate (the number of elements that function correctly / the number of manufactured elements).

Means for solving the problem

  In a pixel peripheral recording type image pickup device in which each pixel is provided with a CCD memory section having a straight line longer than the pixel length or a shape close to a straight line, a large space generated between the end section of the readout vertical CCD and the readout horizontal CCD is utilized. Thus, an area for temporarily storing charges is provided on the side surface of the end section of the readout vertical CCD.

  At the same time, a collision ionization multiplication CCM is provided at the end of the horizontal CCD.

  In addition to a normal forward transfer voltage pattern, a means for generating a reverse transfer voltage pattern is added to the drive voltage generator for transferring charges on the readout vertical CCD.

  As described above, the signal charge of a certain threshold value or more is temporarily stored on the side surface of the read vertical CCD, and only a small signal charge less than that is first multiplied and read by the CCM. Next, the charges on the end section of the readout vertical CCD are evacuated to empty them, and the large signal charges temporarily stored there are returned and read out without CCM amplification.

  As described above, it is possible to add an ultra-high sensitivity and a large dynamic range imaging function to a pixel peripheral recording type imaging device that enables ultra-high-speed video imaging with a simple and reduced structure of the number of metal wires.

  Alternatively, the end section of the readout vertical CCD is provided with a means for charge transfer independently of the upper part thereof, so that the small remaining in the end section without reverse charge transfer. By transferring only the signal charge first, it can be emptied once.

  FIG. 1 shows a device in which a CCM 109 is inserted between the readout horizontal CCD 105 and the readout amplifier 106 of the pixel peripheral recording type imaging device 100 and an image signal temporary storage area 110 is added to the end section 107 of the readout vertical CCD. It is a conceptual diagram. The read amplifier 106 is connected to an AD converter 111 outside the image sensor.

  FIG. 2 shows a system 200 of the photographing apparatus. The system includes a photographing unit (camera unit) 134, a control unit 115, and a monitor 114. In the photographing unit 134, the light 202 incident on the lens 201 forms an image on the pixel peripheral recording type image pickup device 100, and generates an image signal (electrons). The image signal is read from the image sensor 100, amplified, digitized by the AD converter 111, and stored in the buffer memory 112. The control unit 115 includes an image configuration circuit 113 and a photographing unit control device 133. The image construction circuit 113 reads an image signal from the buffer memory 112 in the photographing unit 134, composes it as an image, and outputs it to the monitor 114 at a slow motion of 10 frames / second. The imaging unit control device 133 controls the imaging device 100, the AD converter 111, the buffer memory 112, and the like. In the figure, the image sensor control device 116 in the photographing unit control device 133 is shown separately enlarged. Among them, a CCD drive voltage generator 117 is included.

  The CCD drive voltage generator 117 includes a reverse drive voltage generator 119 of a readout vertical CCD and an image signal temporary storage area 110 (shown in FIG. 1) in addition to the standard drive voltage generator 118 of the pixel peripheral recording type image pickup device. An operating voltage generator 120 is provided.

  FIG. 3 illustrates the end section 107 and the side image signal temporary storage area 110 of the readout vertical CCD of FIG. 1 in more detail. FIG. 4 shows the state of storage of the electrodes 122, 127, 126, potentials, and signal charges in the CC ′ section of FIG.

  The amplification factor of the CCM is 10 times, and the readout noise of the readout amplifier is equivalent to 10 electrons. Therefore, ideally, even one electron is amplified by a factor of 10 and becomes more than the readout noise of the readout amplifier and can be detected.

  The AD converter 111 has 12 bits. That is, the ratio between the minimum detectable unit and the maximum value is 4,096. Since the noise level of the amplifier is 10 electrons, up to about 40,000 signal electrons can be detected.

  This CCD is a four-phase drive. Therefore, in FIG. 3, four electrodes 121, 122, 123, and 124 are mounted on one CCD element 129 in the end section 107 of the readout CCD. In FIG. 1, six CCD elements 129 are connected to constitute the end section 107 of the readout CCD. By sequentially changing the voltage applied to these four electrodes, the signal charge 125 is transferred downward. For this purpose, four types of voltage patterns for four types of electrodes are generated by the standard driving voltage generator 118 shown in the figure.

  Further, in this embodiment, the reverse voltage generator 119 generates four types of reverse voltage patterns for these four types of electrodes. However, since the forward transfer voltage pattern and the reverse feed voltage pattern are both sent by the same metal wiring, it is not necessary to wire a new metal line on the image sensor.

  An electrode 126 is also on the image signal temporary storage area 110, and a gate electrode 127 is between the end section 107 of the readout vertical CCD and the image signal temporary storage area 110. That is, two electrodes 126 and 127 are added in addition to a normal CCD. The operation voltage generator 120 in the image signal temporary storage area in the drive voltage generator 117 shown in FIG. 2 generates two types of voltage patterns for these two types of electrodes.

  The size of the CCD element in the recording area 102 is 3 microns wide and 3.6 microns long (not shown). On top of this, there are four electrodes with a width of 0.9 microns (not shown). The charge transfer capacity is 30,000 electrons. In FIG. 1, since the number of CCD elements in the recording area 102 is 15, continuous 15 images can be continuously overwritten and recorded.

  FIG. 4A to FIG. 4F show the function of the image signal temporary storage area when the signal charge is large.

  First, FIG. 4A shows a state immediately after signal charges are transferred from above. 10V is applied to the electrodes 122 and 123, and 0V is applied to the electrodes 121 and 124. Therefore, the signal charge 125 is stored under the electrodes 122 and 123. Since 0 V is also applied to the electrode 126 and the gate electrode 127 on the image signal temporary storage area 110, the charge has not yet moved to the image signal storage area.

  Next, as indicated by arrows 135 and 136 in FIG. 4B, the electrode 126 on the image signal temporary storage area is raised from 0V to 12V, and the gate electrode 127 is raised to 9V. The difference between the voltage 10V of the electrodes 122 and 123 and the voltage 9V of the electrode 127 is 1V as a potential barrier to the side image signal temporary storage area 110. As indicated by an arrow 137, the charge 132 exceeding the value is transferred to the image signal temporary storage area. At this time, the number of signal electrons of the image signal 130 remaining in the transfer path at the end section of the readout CCD is 3,000.

  For example, when the number of signal electrons is 30,000 which is full of the transfer capacity, 27,000 subtracted 3,000 are temporarily moved and stored in the image signal temporary storage area 110.

  When the original number of signal electrons is 3,000 or less, no signal electrons are stored in the image signal temporary storage area 110. This state is shown in FIG.

  Next, in FIG. 4C, as indicated by an arrow 138, the voltage of the gate electrode 127 is 0V. In this state, 3,000 signal electrons remaining in the transfer path under the electrodes 122 and 123 are read and sent to the horizontal CCD and further sent to the CCM.

  In CCM, the original signal is amplified 10 times. That is, 3,000 signal electrons are amplified to 30,000 on average. Since the number of detectable signal electrons is about 40,000, it can be measured even after amplification by CCM.

  When the 3,000 signal electrons remaining under the electrodes 122 and 123 are transferred to the lower readout horizontal CCD, they are automatically read out as shown by the thick arrow 139 in FIG. The next six signal charges 131 are transferred from the portion immediately above this section and stored in the end section 107 of the readout vertical CCD.

  In order to return the signal electrons 132 once stored in the image signal temporary storage area to the transfer path, it is necessary to once empty the portion of the transfer path below the electrodes 122 and 123. The signal charge 131 falling in the end section 107 of the readout vertical CCD in FIG. 4D is sent back as shown by the thick arrow 140 in FIG. 4E to empty this portion. By performing this operation six times, the state shown in FIG.

  In FIG. 4F, as indicated by an arrow 141, the voltage of the electrode 126 on the image signal temporary storage area returns to 0V. Further, the gate electrode 127 is also returned to 0V. As a result, as indicated by an arrow 142, the signal charge 132 that has once overflowed and stored in the image signal temporary storage area returns to the end section of the readout vertical CCD. Further, the readout is sent from the horizontal CCD to the CCM. In this case, the CCM is operated at the same voltage as a normal readout horizontal CCD, so secondary electrons are not generated and there is no amplification effect. Therefore, only 27,000 signal electrons can enter the read amplifier. This is a read amplifier capacity of 40,000 electrons or less.

  Thereafter, the state returns to the state of FIG. 4A and the above operation is repeated.

  As described above, since a minimum of one signal electron can be detected from a maximum of 30,000, a large dynamic range close to 15 bits (about 32,768) and two ultra-high sensitivity of a minimum of one signal electron can be detected. Ultra-high-speed continuous video shooting of over 1 million frames per second has become possible with the newly added high-performance pixel-peripheral recording image sensor.

  If there is no function for temporarily storing large signal electrons in the image signal temporary storage area, both small signal electrons and large signal electrons are uniformly amplified, and the original 30,000 signal electrons are amplified to a maximum of 300,000. As a result, it overflows far beyond the capacity of 40,000 electrons of the read amplifier, and the dynamic range becomes small.

  On the other hand, when the CCM function is not activated, overflow in the readout amplifier does not occur, but small signal electrons of 10 electrons or less are not detected, and ultra-high sensitivity imaging cannot be performed.

  FIG. 5 shows the case where the original number of signal electrons is 3,000 or less. Since only the signal electrons do not overflow in the side image signal temporary storage area, the rest is the same as in FIG.

  As shown in FIG. 1, in the pixel peripheral recording type imaging device, since there is a large space at the end of the readout vertical CCD and the junction of the readout horizontal CCD, it is easy to incorporate such an image signal temporary storage area. is there. Therefore, there are only two types of electrodes that need to be added, that is, the electrode 126 and the gate electrode 127 on the image signal storage region. Further, only two metal wires (not shown) for sending voltages for operating them are added.

  Since the voltage pattern for reverse running is generated by the CCD drive voltage generator in the control device outside the image sensor, it is necessary to create a special structure for the inside of the image sensor having a complicated and fine structure. Absent.

  As described above, in this embodiment, by adding an extremely simple structure to the image sensor, a high function of a wide dynamic range is newly added in addition to the ultra high speed and the ultra high sensitivity.

  The difference from the first embodiment is that the end section 107 of the readout vertical CCD can be transferred independently of the section above that of the readout CCD. That is, four metal lines (not shown) are wired in the image pickup device in order to transfer signal charges in the end section 107 of the readout vertical CCD. In addition, a transfer voltage generator (not shown) for generating this voltage pattern is formed in a control device outside the image sensor.

  When there are 3,000 or more image signals, the image signal once overflows into the image signal storage area (state (c) in FIG. 4), and then 3,000 charges are read and transferred to the horizontal CCD. At this time, only the CCD at the end section of the readout vertical CCD is operated independently of the section above it and is emptied. Thereby, the image signal stored above it does not go down. Therefore, there is no need to reverse feed. That is, as compared with the first embodiment, the process of FIG. 4E can be omitted, and the operation is simplified.

  However, the image signal storage region electrode voltage and the gate electrode voltage are combined to generate a total of six types of transfer voltage patterns, and six types of metal lines for feeding them into the image sensor are used as the image sensor. Built in.

  Super high speed shooting. In ultra-high-speed shooting, the amount of light is insufficient and high sensitivity is required, and the brightness difference between a bright part and a dark part around it is large, and high dynamic range shooting is often required. Typical observation objects include explosion phenomena, combustion phenomena in engines, laser ablation processing, high-speed phenomena in cells (fluorescence observation), and the like.

Conceptual diagram of a pixel peripheral recording type image pickup device according to the present invention System diagram of the imaging device Detailed view of end section of readout vertical CCD Storage state of electrodes, potentials and signal charges in section CC 'in FIG. State of the end section of the readout vertical CCD when the number of signal charges is 3,000 or less

Explanation of symbols

100 pixel peripheral recording type image pickup device 101 photodiode 102 linear CCD memory of image signal recording area (consisting of 15 elements)
103 Drain 104 Read vertical CCD
105 Readout horizontal CCD
106 Readout amplifier 107 Readout end section of vertical CCD (consisting of 6 elements)
108 Wasted space between the readout vertical CCD and readout horizontal CCD 109 CCM
110 Image signal temporary storage area (consisting of 6 elements)
111 AD converter 112 Buffer memory 113 Image configuration circuit 114 Monitor 115 Control unit 116 Image sensor control device 117 CCD drive voltage generator 118 Standard drive voltage generator 119 Read vertical CCD reverse voltage generator 120 Operation of image signal temporary storage area Voltage generators 121, 122, 123, 124 Voltage 125 on the CCD element in the end section of the readout CCD Signal charge 126 Electrode 127 on the image signal temporary storage area Gate electrode 129 One CCD element in the end section of the readout vertical CCD 130 Signal charge 131 remaining in the end section of the readout CCD 131 Signal charge falling from above in the end section of the readout CCD 132 Signal charge overflowing the temporary storage area of the image signal 133 Imaging unit controller 134 Imaging units 135 to 135 142 Potential change System 201 lens 202 incident light arrows 200 imaging apparatus according to the movement of charges

Claims (4)

A pixel array of two or more rows and two or more columns, and each pixel includes a straight line-shaped CCD type memory unit longer than the pixel length or more than one row extending in the pixel row direction outside the pixel array An imaging device comprising a readout horizontal CCD, a collision ionization multiplying CCD, and an output amplifier,
Each pixel column is provided with one column of read vertical CCDs extending in the column direction of pixels between the pixel array and the readout horizontal CCD, and a region for temporarily storing charges is provided on the side surface. Image sensor.   3. The image pickup device according to claim 1, further comprising means for transferring charges on the readout vertical CCD in two directions, ie, a charge reading direction and a reverse direction.   3. The read vertical CCD comprising: a portion having a region for temporarily storing charges on a side surface thereof; and a portion on the upper side thereof, each having means for transferring charges independently. Image sensor   An imaging device comprising the imaging device according to claim 1, claim 2, or claim 3.

Images