RU1140651C - Device for processing information - Google Patents

Description

The invention relates to semiconductor technology and can be used in transmitting television devices for routine processing of video signals.

A device for processing information on a charge-coupled device (CCD) is known, which consists of a storage section, a storage section of an output shift register with information input devices and an output shift register with an information output device.

A drawback of this matrix design is the inability to isolate video signals at the output devices for further current processing.

The closest technical solution is an information processor on a CCD, consisting of three input registers with bandwidth) in two mutually perpendicular directions with a common input device and three matrix sections with similar output shift registers. There are 8 three elements of which contain non-destructive reading devices information based on floating gates. Such a design makes it possible to isolate a 3 X 3 matrix of signals at the output devices, but it is not possible to process the signals on a chip by using the Zobel transform to select the edges of the image, which reduces the functionality of the device.

The aim of the invention is to expand the functionality of the device by implementing the transformation of Zobel on a chip.

The goal is achieved in that in the information processing device on the CCD. consisting of three input shift registers with a capacity in two mutually perpendicular directions with a common input device and three matrix sections with similar design shift output registers, and the three elements of which contain the main devices for non-destructive reading of information based on floating gates, additionally contains in the first and third output shift registers, in the first and third elements with devices for non-destructive reading of information, one additional Nome identichmomu nondestructive readout device information; there are also two schemes for determining the difference modulus and a summing gate, the gates of the main devices for non-destructive reading of information connected to each other in the first and third output shift register and connected to the inputs of the first circuit for determining the difference module, and the gates in the first and third elements of the first output shift register non-destructive reading devices, information is connected respectively to the gates in the first and third elements of the second output shift register of the main devices non-destructive information reading TVs and, respectively, with gates in the first and third elements of the third output shift register of additional non-destructive information reading devices and are connected to the inputs of the triple difference module determination circuit. In FIG. 1 is a schematic diagram of a device: FIGS. 2 and 3 are a diagram for determining a difference module: FIG. 4 is a timing diagram of a control voltage.

The information processing device on the CCD consists of three input shift registers 1,2,3 with a common input device 4. The input device is a diffusion region and two gates. Input shift registers have bandwidth in two mutually perpendicular directions and are identical in design and size. Matrix sections 5i6,7 are equipped with output shift registers 8,9,10, identical in design to the input ones. In the three elements of each output shift register there are devices for non-destructive reading of information based on floating gates 11-22, In the output shift registers 8IO there are 51 solid-state devices for non-destructive reading of information with; gates 14, 15 and 21, 22, respectively. The main gates 11, 12, 13 are connected to each other (point K) and are connected to the input of the primary circuit for determining the difference modulus. The main gates 18, 19, 20 are interconnected (point N) and connected to the input P of the first difference module determination circuit. Additional gates 21.14 are connected to the main gate 16 (point L) and connected to the input S of the second difference module determination circuit. Additional gates 15.22 are connected to the main 17 (point M) and connected to the input R of the second difference module determination circuit. Structurally, the circuit for determining the difference modulus is an input diffusion region (input 511406 diode) 23 and transfer gates 24, 25, 26. moreover, the gate 24 is connected to the gate 27 and the gate 26 to the gate (X) m 28. The second module determination circuit difference has a logical design. The gates 26,27,28,30 5 are in charge communication with the summing gate 31. The device is made using a three-layer polysilicon technology. The regions limiting the transfer channels are formed by implantation of a mouse, | 0 diffusion regions of the input and output devices by ion doping of the boron; silicon dioxide is a 0.12 micron thick dielectric coating. In the time diagram of Fig. 4, F1P, ° F2P and FZR are clock pulses which are supplied to the register phase electrode, FS, F2M. L3M - on the phase electrodes of the matrix; Д1Р - on the input diode of the input device 4):; Д2Р - on the input diode 23 of the difference modulus determination circuits. F1P on the summing shutter 31 (figure 1). Section A at the timing of the clock diagram provides input and reading of 35 information signals: section B - vertical charge transfer. Impulse amplitudes The symmetry and resolution of the order of V, the lower level is 13 V, the upper is ZV. The pulse amplitudes D IP and D2P are 3-4 V. Consequently, we begin the matrix of information signals of three consecutive lines in 2a 2b 2c ri i 3f. 2 o 2h 2 g. The introduction of additional identical .. devices for read-only information reading (Fig. 1) 14, 15, 21, 22 in the first and third output shift registers makes it possible to separate the information charge into two equal parts. As a result, at the gate pax 14 and It, the signal a is released, at the gate; pax 13 and 15 - b, on the shutters 21 and 18 - e and nazatvorami22. Since in the first shift register the main gates 11, 12 and TK are interconnected (point K), then with the total change in potential at point K when the infeedment charges for floating gates are equal to Ai (a + 2b + c): A similar connection between each other (point M the main gates 18, 19 and 20 in $ 5 the third shift register allows you to receive the signal & a (9 + 2h + 1), Points K and N are connected to the inputs O and P of the first circuit for determining the difference modulus; as a result of 516 we get at the circuit output , signal Si 1 Ai - Zie I. A similar connection of additional shutter 14 and 21 with the main shutter 16 (point L) allows you to select the signal Дз - (а + 2d д), while connecting additional shutters 15 and 22 with the main 17 (point М) -signal Д) - (), Points m and L the difference modulus is connected to the inputs R to S of the second circuit; as a result, at the output of the circuit, we obtain the signal S2 I Дз - А «1-Summing gate 31 serves to obtain the sum S (e} St + 82 | d - (. j Дз - D I (a + 2b + c) - (g + 2h + l) (a + 2d-bHHc + 2f + I). which is an implementation of the Zobel transform used to emphasize outlines in an image. The organization of the device is three phase. The device operates as follows. Consider entering three consecutive information lines A. B and C into an information processing device. The signal corresponding to line A is fed to a common input device. Since all input registers are identical, line A fills equally each input register. Let us denote the contents of registers 1.2,3 - At, A2, Az, respectively. Next, the information charge At, A2, Az is shifted through the matrix section 5 until the register B is filled with the A1 line. The input of the B line occurs synchronously with the reading of the At line. We introduce the corresponding notation B 1, B2, Vz. I tact the phase voltage at the electrodes of Matpi) 4Hbix sections and registers, shift the information until the string Aa is filled (line 9 and line B of register 8. The lines A2 and Bi are registered by registers 9 and 8 synchronously with the introduction of a new information line C. Further shift The information leads to filling in the staaz register 10, line Bz - register 9 “line Ci-register 8. Reading of signals of three consecutive lines is carried out on output devices of non-destructive reading Formations based on floating gates 11-22. the gate is set using the MOS transistors. We denote the signal charges of three consecutive bits of the registers 8. 9 and 10 through 2a, 2b, 2s. 2s1, 2e, 2f and 2g, 2h, 21, respectively. corresponding to information charge a, on shutters

13.15 - s, on the gates 18. 21 and 20. 22 respectively d and I. and on the gates 12.16.17 and 19 - 2b. 26.21 and 2h, respectively, the expression (a + 2b + c) will correspond to a change in potential at point K. 8 point M (d + 2h +1). at the point L - (a d). at point M (c + 2f + O-So, the signal fjH-2b + c is fed to the input O of the Definition circuit of the difference module), sig | 4al is input to the input P (d + 2h +%, the signal will be output 1 {a + 2b c) - (d + 2h + 1}. The signal Si is generated as follows: Let us consider the case of Ai (Figs. 1 and 2). The si-phase signals & q and respectively .8 the same time, the injection of the charge from the diode 23. takes place. Sn1 “keny potential on the diode 23 the upper level of the potential filled with injected healthy building under the gates 24. 26 will be determined by the value of Dg. And the lower level under the gate rum 26 aolichinoy D2. By reducing the potential at the summing gate 31 is drained charge,

corresponding (Ai-4e). to the potential gate 31. In the case of Ai Dg, a charge (Zte-AI) is formed on the gate 27 of the first & nd scheme for determining the difference modulus, while the charge formed at the first output is 0.

Similarly, the second difference module determination circuit works, i.e. in the depletion region, under the summing electrode, a charge S (e) Sj + 82} At- / 5gl + 1D} - Ajl is formed. It is an implementation of the Zobel transform on the CCD, so that it is possible to select the contours in the image in real time;

20 (56) Charging devices. / Ed. M. Khouaz. Moscow: Energoizdat, 1981. p. 314-316.

Copyright copyright USSR No. 93675 &. class and 01 L 27 / 10.1982.

5-1

g

at a, {inZk e} f

S

 behind

zg

V

g

fue. 1