JPH04180465A - Semiconductor integrated circuit for picture signal correction - Google Patents

Abstract

PURPOSE:To reduce the cost and to simplify the manufacture by using a dark correction circuit and a shading correction circuit whose hardware configuration is identical and forming a picture correction circuit not requiring a divider. CONSTITUTION:When a picture data Vi is inputted to a dark correction circuit 5, a subtractor 11 is used to calculate a difference Vi-Bi with a black reference data Bi stored in a line memory 13 for data correction and a transformation circuit 31 applies logarithmic transformation to send log(Vi-Bi) to a shading correction circuit 6, stored in a line memory 23 and used for a reference of shading correction. A subtractor 32 of the correction circuit 6 takes a difference between log(Vi-Bi) and log(Wi-Bi) and a transformation circuit 33 applies inverse logarithmic transformation to obtain (Vi-Bi)/(Wi-Bi) to apply shading correction. Since the dark correction circuit and the shading correction circuit are identical in terms of the hardware, the circuit is simply and inexpensively manufactured.

Description

[Detailed description of the invention] [Industrial application field]

The present invention relates to an image signal correction semiconductor integrated circuit that performs dark correction or shading correction on an image signal.

[Conventional technology]

When an image of a document is read by an image sensor, the output of the image sensor is subjected to corrections such as dark correction and shading correction, and then outputted as an image signal. FIG. 4 shows such an image reading device. In FIG. 4, 1 is an image, 2 is a light source, 3 is an image sensor, 4 is an AD conversion circuit, 5 is a dark correction circuit, 6 is a shading correction circuit, 7 is an output terminal. The dark correction circuit 5 and the shading correction circuit 6 are configured as semiconductor integrated circuits. The image sensor 3 reads the image by receiving the reflected light of the image 1 illuminated by the light source 2, and its output is AD
The conversion circuit 4 converts it into a digital quantity. Then, after being corrected by the dark correction circuit 5 and the shading correction circuit 6, the signal is outputted from the output terminal 7 as an image signal. FIG. 5 shows a conventional semiconductor integrated circuit for image signal correction. The symbols correspond to those in FIG. 4, and 8 is a data input terminal;
9 is a clock input terminal, 10 is a D flip-flop, 1
1 is a subtracter, 12 is a D flip-flop, 13 is a line memory, 14 is a timing controller, 15 is a start signal line, 16 is a write command signal line, 17 is an address signal line, 18 is a write enable signal line, 19 is a Output data enable signal line, 20 is a D flip-flop, 2
1 is a divider, 22 is a D flip-flop, 23 is a line memory, 24 is a timing controller, 25 is a start signal line, 26 is a write command signal line, 27 is an address signal line, 28 is a write enable signal line, 29 is an output This is a data enable signal line. D flip flop! 0.12.20.22 is provided to synchronize the data to the clock when it is sent to the next stage. Line memory 13.23 is a memory for storing data for one line (row) of an image, and R
AM (Random Access Memory) is used. Timing controllers 14.24 are for timing the operation of each correction circuit. Note that AD means address, WE means write enable, and DE means data enable. Next, the operation of the circuit will be explained. Prior to correction, the following preparations are made. First, look at image 1 in Figure 4! ! L, the reference plane, and turn off the light source 2. The output of the image sensor 3 at this time is AD
The data is inputted to the data input terminal 8 in FIG. 5 through the conversion circuit IM4. The line memory 13 uses this as black reference data B.
will be written to. Writing is performed according to the AD times signal and WE times signal from the timing controller [4]. This is used as a reference value for dark correction. FIG. 6 is a diagram for explaining dark correction, in which the horizontal axis represents the position in the line direction (longitudinal direction of the image sensor 3), and the vertical axis represents the output of the image sensor 3. FIG. 6(a) shows black reference data B. Ideally, the output of the image sensor 3 when reading the black reference plane is zero, but in reality it is not zero and some output is output as shown. This is, so to speak, an offset amount. Next, image 1 is used as a white reference plane, and light source 2 is turned on. The output of the image sensor 3 at this time is input to the data input terminal 8 of FIG. 5 via the AD conversion circuit 4. Let this be white reference data W, and the difference W from the black reference data B written in the line memory 13 by the subtracter 11 is -
Take B. This difference is sent to the shading correction circuit 6 and written into the line memory 23, and is used as a reference value for shading correction. Let image 1 be the image to be read, and the image data to the data input terminal 8 when read by the image sensor 3 with the light source 2 turned on is v3. Then, in the subtracter 11, ■1 and black reference data B! The difference v, -Bi is taken. Taking this difference is dark correction, and in FIG. 6 (b), it is correction that subtracts black reference data B, which is an offset, from . The dark-corrected image data V, -B is divided by W, -B stored in the line memory 23 by the divider 21 of the shading correction circuit 6. The resulting V, -B. W, -B. is the image data that has undergone shading correction. This is output from the output terminal 7.

[Question H that the invention attempts to solve]

However, the conventional semiconductor integrated circuit for image signal correction described above has the following problems. The first problem is that since the dark correction circuit and the shading correction circuit have different configurations, they must be manufactured separately, which is troublesome and increases costs. The second problem is that the shading correction circuit must include a divider, but the divider requires a large number of transistors. An object of the present invention is to solve the above-mentioned problems.

[Means to solve the problem]

In order to solve the above problems, the semiconductor integrated circuit for image signal correction of the present invention includes a line memory that stores a reference value for correction, a subtracter that subtracts the reference value from input data, and a select signal that outputs the output of the subtracter. The system is equipped with a LOG/inverse LOG conversion circuit that performs LOG conversion or inverse LOG conversion according to an instruction.

[For production]

The dark correction circuit and the shading correction circuit, which are semiconductor integrated circuits for image signal correction, have exactly the same hardware configuration, so it is possible to reduce the manufacturing effort and reduce costs. Become. Further, since the image signal correction semiconductor integrated circuit does not include a divider, the number of transistors is reduced.

【Example】

Embodiments of the present invention will be described in detail below with reference to the drawings. FIG. 1 shows a semiconductor integrated circuit for image signal correction according to the present invention. The symbols correspond to those in Figure 5, and 31 is LOG/inverted L.
OG conversion circuit, 32 is a subtracter, 33 is LOG/inverse LOG
A conversion circuit, 34 is a select signal input terminal, and 35 is an inverter. The LOG/inverse LOG conversion circuit is a circuit that can be used both as an LOG conversion circuit and as an inverse LOG conversion circuit, and switching thereof is performed by a select signal. The LOG conversion circuit is a circuit that converts input into a LOG value. Figure 2 is a diagram showing LOG conversion, where the horizontal axis is the input,
The N axis indicates the output.For example, when r7° is input, rLog7' is output. The inverse LOG conversion circuit is a circuit that converts an input LOG value into a value without LOG. FIG. 3 is a diagram showing inverse LOG conversion, where the horizontal axis represents input and the vertical axis represents output. For example,
When "rLog7" is input, "7" is output. In the present invention, the dark correction circuit 5 and the shading correction circuit 6 have exactly the same hardware configuration. However, in the dark correction circuit 5, the LOG/inverse LOG conversion circuit 31 is
The only difference is that the LOG/inverse LOG conversion circuit 33 in the shading correction circuit 6 is operated as a G conversion circuit, and the LOG/inverse LOG conversion circuit 33 is operated as an inverse LOG conversion circuit. Next, the operation will be explained. Black reference data B8 is stored in the line memory 13 in the same manner as in the prior art. This is used as a reference value for dark correction. Next, similarly to the conventional method, the output of the image sensor 3 when the image 1 is a white reference plane is sent to the dark correction circuit 5 via the AD conversion circuit 4 shown in FIG. 4, that is, to the data input terminal 8 shown in FIG. Enter. Let this be white reference data W, and subtracter 1
1, the difference W, -B1 from the black reference data B stored in the line memory 13 is taken. This difference is LOG/reverse LO
The G conversion circuit 31 performs LOG conversion and log (WH-B
H) is obtained. This is sent to the shading correction circuit 6,
It is stored in the line memory 23. This is used as a reference value for shading correction. Let image 1 be the image to be read, and let the image data to the data input terminal 8 when read by the image sensor 3 with the light source 2 turned on be V. Then, the subtracter 1
1, the difference V, -B, between ■ and the black reference data B is taken, dark correction is performed, and further LOG conversion is performed in the inverse LOG conversion circuit 31, and log (VH-BH
). In the shading correction circuit 6, the input log
The difference between (V; -B*) and log (WH-B;) stored in the line memory 23 is taken. l
og (VH-BH) -1og (WHB; ) The above result obtained by the subtracter 32 is subjected to inverse LOG conversion in the LOG/inverse LOG conversion circuit 33, and becomes V, -B. W, -B. becomes. As a result, shading-corrected image data similar to the conventional one can be obtained. Note that when the inverter 35 receives a select signal from the select signal input terminal 34 that causes the LOG/inverse LOG conversion circuit 31 to function as an inverse LOG conversion circuit, the inverter 35 uses that signal to cause the LOG/inverse LOG conversion circuit 33 to function as an inverse LOG conversion circuit. It is inserted to create a select signal that works as a select signal. Dark correction circuit 5 and shading correction surface ll of the present invention
Since the hardware configuration is the same as ll6, the same semiconductor integrated circuit can be used, which requires less effort and costs less than creating a different circuit as in the past.

【Effect of the invention】

As described above, the semiconductor integrated circuit for image signal correction of the present invention provides the following effects. (2) Since the dark correction circuit and the shading correction circuit, which are semiconductor integrated circuits for image signal correction, have exactly the same hardware configuration, manufacturing time and effort can be reduced and costs can be reduced. (2) The semiconductor integrated circuit for image signal correction does not include a divider, so the number of transistors is reduced.

[Brief explanation of drawings]

Fig. 1... Semiconductor integrated circuit for image signal correction of the present invention Fig. 2... Diagram showing LOG conversion Fig. 3... Fig. 4 showing inverse LOG conversion... Image reading device No. 5 Figure: Conventional semiconductor integrated circuit for image signal correction No. 6
Figure... In the figure explaining dark correction, ] is an image, 2 is a light source, 3 is an image sensor, 4 is an AD conversion circuit, 5 is a dark correction circuit, 6 is a shading correction circuit, 7 is an output terminal, 8 is a data input terminal, 9 is a clock input terminal, 10 is a D flip tab,
11 is a subtracter, 12 is a D flip-flop, 13! ,t
Line memory, 14 is a timing controller, 15 is a start signal line, 1G is a write command signal line, 17 is an address signal line, 18 is a write enable signal line, ]9 is an output data enable signal line, 20 is a D flip hook tab, 21 is a divider, 22 is a D flip-flop, 23 is a line memory, 24 is a timing controller, 25 is a start signal line, 26 is a write command signal line, 27 is an address signal line, 28 is a write enable signal line, 29 is an output Data enable signal line, 31 is LOG/reverse LOG
32 is a subtracter, 33 is a LOG/inverse LOG conversion circuit, 34 is a select signal input terminal, and 35 is an inverter. Patent Applicant Fuji Xerox Co., Ltd. Representative Patent Attorney Tomio Honjo Figure 4 → Lie/Direction Position (A) No. 6 First Lie/Direction Position (B) Figure

Claims (1)

[Claims] It includes a line memory that stores a reference value for correction, a subtracter that subtracts the reference value from input data, and an LOG/inverse LOG conversion circuit that converts the output of the subtracter into LOG or inverse LOG according to the instruction of a select signal. A semiconductor integrated circuit for image signal correction, characterized in that: