JPH0448033B2 - - Google Patents

Description

[Detailed description of the invention]

INDUSTRIAL APPLICATION FIELD The present invention relates to an image signal detection and correction device for a color video camera or the like using a solid-state image sensor. Conventional configurations and their problems In recent years, color video cameras using solid-state image sensors have made remarkable progress, and their performance has increased and they have become more multifunctional. Among these, improving basic image quality performance is the most important element. Currently, in order to improve this image quality, it is necessary to remove false signals that appear on the screen, and signal correction in each part of the signal processing device is being considered. FIG. 1 shows a method for sequentially extracting color difference signals from pixel columns in a solid-state image sensor arranged in the scanning direction so that different color signals are alternately repeated for each pixel as shown in FIG. 1a. A _i and B _i (i is a natural number) in the figure represent each pixel signal level. That is,
Create sets in units of two pixels in the scanning direction, take the difference between each set, and obtain color difference signals A ₁ −B ₁ , A ₂ −B ₁ , A ₂ −B ₂ , A ₃
−B ₂ ... is obtained point-sequentially. FIGS. 1b and 1c show how the color difference signal changes when a boundary edge 11 of an object exists in a solid-state image sensor having such a pixel array. In order to obtain correct color difference signals, adjacent pixel pairs (A _i , B _i ) or (A _i+1 , B _i ) must receive light having the same luminance and color components. However, the boundary edge 11 of the subject is a part where the brightness component or color component changes rapidly.
As can be seen from FIG. 1c, at the border, different luminance or color components are mixed into the set of pixels from which the color difference signals are obtained, making it impossible to obtain a correct color difference signal, resulting in a so-called false signal 12. Therefore, the boundary edge portion 11 is colored differently from the actual color, and the image quality is significantly degraded. Conventionally, in order to remove this false signal, the color difference signal has been reduced around the boundary edge. In this case, however, the image quality is improved to some extent at the boundary edges of completely achromatic subjects, but at the boundary edges of chromatic subjects, even more unnatural contours occur, which actually deteriorates the image quality. It has also been considered to insert an optical low-pass filter to alleviate edge changes and eliminate false signals, but this poses the problem of significantly lowering the resolution. Purpose of the invention The present invention solves the above-mentioned conventional problems.
If the image signal data contains false signals,
It is an object of the present invention to provide an image signal detection and correction device that removes false signals without reducing resolution by replacing them with other image signal data. Configuration of the Invention The image signal detection and correction device of the present invention obtains level difference data or a level ratio value between an h-th image signal and an (h+1)-th image signal generated point-sequentially from a solid-state image sensor, and A function that compares the level difference data or level ratio value with preset reference data, and replaces the image signal data generated point-sequentially from the solid-state image sensor with other image signal data according to the comparison result. This makes it possible to remove false signals and improve image quality. DESCRIPTION OF EMBODIMENTS An embodiment of the present invention will be described below based on the drawings. FIGS. 2a and 2b are signal processing block diagrams of the image signal detection and correction apparatus of the present invention. Hereinafter, for ease of explanation, it is assumed that the pixel array of the solid-state image sensor 21 is as shown in FIG. 1a. In FIGS. 2a and 2b, 21 is a solid-state image sensor;
22 is a color signal processing device, 23 is an image signal processing device, 24 is an image signal level change detection device, 25 is a comparison/judgment device, 26 is a color signal data replacement device, 27
is a control signal. The difference between FIG. 2 a and b is that the color signal processing device 22
The difference is whether it is located before or after the image signal replacement device 26. That is, in the case of FIG. 2a, the color signal processing device 22 performs color signal processing on the pixel signals sequentially taken out from the solid-state image sensor 21 to obtain color difference signals, and then the color difference signals are used as images. The signal replacement device 26 replaces the control signal 27 with the control signal 27 . At this time, the image signal corresponds to the color difference signal. On the other hand, in the case of FIG. 2b, the image signals sequentially taken out from the solid-state image sensor 21 are replaced by the control signal 27, and then the color signal processing device 22 obtains all correct color difference signals. At this time, the image signal corresponds to the pixel signal. Figures 3a and b are respectively Figures 2a and b of the present invention.
This shows the operating principle of the block diagram. Third
In the case of FIG. a, as described above, the color difference signals A ₁ −B ₁ , A ₂ are obtained point-sequentially from the pixel signal string of the solid-state image sensor 21.
−B ₁ ... is obtained. However, when the signal change 31 of the object occurs, for example, between the pixel A ₂ and the pixel signal B ₂ , the color difference signal A ₂ -B ₂ becomes a false signal. Therefore, by using the color difference signal A ₃ -B ₂ immediately after this false signal A ₂ -B ₂ instead of this false signal, the false signal can be removed. At this time, the false signal A ₂ −
If B ₂ is the hth color difference signal, then h+1
This means that the color difference signal A ₃ −B ₂ is used instead. In the case of FIG. 3b, when obtaining a color difference signal, a pair is always made of a pixel A _i and a pixel signal B _i (i is a natural number), and A _i -B _i is used as the color difference signal. Therefore, a color false signal occurs when the signal change 32 of the object is between the pixel A _i and the pixel B _i . At this time, if the pixel signal replacement operation is not performed,
The color difference signal A _i −B _i becomes a false signal. Therefore, if the pixel signal A i ₊₁ is used instead of the pixel A _i , the color difference signal A _i+1 −B _i will be obtained instead of the false signal A _i −B _i , so the false signal can be removed. In FIG. 3b, the case i=2 is shown. At this time, if the false signal A _i −B _i is the hth color difference signal, it is replaced with the h+1th color difference signal A _i+1 −B _i . Next, an example of the configuration of each block will be explained. FIG. 4 is a configuration diagram of the color signal processing device 22 of the present invention. The latch circuit 41 and the subtracter 42 provide the latch circuit 43 with color difference signals A ₁ −B ₁ , A ₂ −
B ₁ , ..., A _i -B _i , A _i+1 -B _i , A _i+1 -B _i+1 , ... are input. If the frequency of the sampling clock 44 of the latch circuit 43 is the same as the frequency of the sampling clock from which the pixel signal is taken out, the color difference signal is output as is. Furthermore, if the frequency of the sampling clock 44 is set to 1/2 of the sampling clock frequency for extracting pixel signals and the phase is adjusted appropriately, then A ₁ -B ₁ , A ₂ -B ₂ , ..., A _i -B _i ... color difference signals are output. Therefore, by determining the frequency and phase of the sampling clock 44 of the latch circuit 43 according to the block diagrams in FIGS. 2a and 2b, an appropriate color difference signal can be obtained. FIG. 5 is a configuration diagram of the image signal processing device 23 of the present invention. The image signal is the sum of adjacent pixel signals that are sequentially extracted. That is, the image signal is A _i +
B ₁ , A ₂ +B ₁ ,…, A _i +B _i , A _i+1 +B _i , A ₁ +1+
B _i+1 , ... are obtained sequentially. By passing the pixel signal through a simple low-pass filter composed of a latch circuit 51 and an adder 52, an averaged signal gain is created. Therefore, even if pixel signal level A is higher than pixel signal level B for a red subject, for example, the sum of these two levels is created and averaged, so an image signal of a certain constant level is always obtained. The signal level does not vary for each pixel signal. FIGS. 6a and 6b are block diagrams of the image signal level change detection device 24 of the present invention. In FIG. 6a, the level difference between the image signals successively produced by the delay circuit 61 and the subtracter 62 is taken as the value of the change in these signal levels. That is, when an image signal is G _i and these are obtained point-sequentially as G ₁ , G ₂ , ..., G _i-1 , G _i , G _i+1 ..., the image signal level change detection device 24 detects G ₁ −G ₂ , G ₂ −G ₃ …, G _i-1 −G _i , G _i −G _i+1 ,
...The level difference data of the image signal is created point-sequentially. In FIG. 6b, the image signal level change is expressed as the ratio of the image signals. In this case, the delay circuit 6
From the image signal level change detection device 24 consisting of 3 and a divider 64, G ₁ /G ₂・G ₂ /G ₃ . . .
The image signal ratio values are obtained as G _i-1 /G _i , G _i /G _i+1 . . . The level difference data or ratio value of the image signal created in this way is used as replacement determination signal data. Level difference data of image signals can be easily obtained using a subtracter. Further, since the signal level change is expressed logarithmically, the value of the ratio of the image signal matches well with the characteristics of the human eye. FIG. 7 is a configuration diagram of the comparison/judgment device 25 of the present invention. In FIG. 7a, image signal level change data is input into two comparators 71 and 72, and control signals 73 and 74 are obtained by comparison with reference data M and reference data N, respectively. Standard data M
is data larger than the reference data N. The reference data M corresponds to the first value or the third value described in the claims, and the reference data N corresponds to the second value or the fourth value described in the claims. When the image signal level change data is larger than the reference data M or smaller than the reference data N, the control signal 73 or 74 is output, respectively.
is output. That is, when the image signal level change data does not exist within the range of the reference data M and the reference data N, the control signal is generated. These two control signals 73,7
Of course, 4 are not output at the same time. When the image signal level change data is image signal level difference data, a configuration of the comparison and determination device as shown in FIG. 7b is also considered. In this case, after obtaining the absolute value of the image signal level change data from the absolute value detection circuit 75, the data is compared with the reference data L by the comparator 76, and when the absolute value is larger than the reference data L, the control signal 79 is Output.
Further, the data sign discrimination circuit 77 determines that the control signal 7 is detected when the image signal level change data is positive.
8 is output, so it can be handled in the same way as the control signals 73 and 74 in FIG. 7a. In FIG. 7b, only one comparator is required, which simplifies the configuration.
In this case, the absolute value of the reference data L corresponds to the first value recited in the claims, and the value with a negative sign of the absolute value of the reference data L corresponds to the second value recited in the claims. corresponds to FIG. 8 is a block diagram of the image signal replacement device 26 of the present invention. As described above, the image signal is a color difference signal obtained from the color signal processing device 22 in the case of FIG. 2a, and a pixel signal from the solid-state image sensor 21 in the case of FIG. 2b. The case where the image signal is a color difference signal will be explained below. When neither of the control signals 73 and 74 is output, the image signal 85 that has passed through the delay circuit 81 is output from the data switching circuit 83. That is, the generation timing of the control signals 73 and 74 and the flow timing of the color difference signal are adjusted around the timing at which the color difference signal data 85 enters the data switching circuit 83. Therefore, by outputting the control signals 73 and 74, the data switching circuit 83 outputs the image signal 86 that does not pass through the delay circuit 182, the image signal 84 that passes through the delay circuits 81 and 82, and the image signal that passes through the delay circuit 81 as described above. 85 are selected. The results are shown in Table 1. Here, "L" and "H" indicate cases in which the control signals 73 and 74 are not output and cases in which they are output.

[Table] As can be seen from Table 1, the switching of image signals differs depending on whether the change in the subject is from a small signal level to a large signal level or vice versa. That is, if the image signal 85 is the h-th signal, the image signals 84 and 86 are h-1 and h+1-th signals, respectively. and,
When the change in the subject is from a small signal level to a large signal level, the h-th image signal is h
Conversely, when the change in the subject is from a large signal level to a small signal level, the h-th image signal is replaced by the h+1-th image signal. The configuration and operating principles of each block have been explained above based on the overall block diagram of the image signal detection and correction device. Note that the image signal generated by the image signal processing device 23 is not an averaged signal of each pixel group, and the pixel signal may be used as it is. At this time, the image signal level change detection device 24 and the comparison/judgment device 25 detect signal level changes between pixel signals of the same type closest to each other in the scanning direction, create a logical product of these change detection signals, and use this as a control signal. That's why. FIG. 9 shows the operation of the image signal level change detection device and comparison/judgment device when pixel signals are used as image signals. Difference signals obtained from the pixel row A _i (A ₁ −A ₂ , A ₂ −A ₃ , ...A _i+1 −A _i , A _i −
A _i+1 …), and the difference signal obtained from the pixel row B _i (B ₁ −B ₂ , B ₂ −B ₃ , …B _i-1 −B _i , B _i −B _i+1 …)
are as shown in FIG. 9c and d, respectively. Therefore, if we take the logical product of these difference signals, we get the ninth
A signal change as shown in Figure e is obtained, and the position 991 of the signal change of the subject can be accurately detected. FIG. 10 is a block diagram of a pixel signal level change detection device and a comparison/judgment device for obtaining the operation shown in FIG. 9. Sampling clock A and sampling clock B input to latch circuits 101 and 102, respectively, have frequencies that are 1/2 of the sampling clock of the pixel signal, and their phases are opposite to each other. Therefore, the pixel signal columns are separated, and the signals of pixel column A and pixel column B are input to image signal level change detection devices 103 and 104, respectively. Then, the signals of each pixel column are separately processed by comparison/judgment devices 105 and 106, and the AND circuit 10
Control signals 109, 11 via 7, 108
0 is lost. Standard data AM and standard data AN are the maximum and minimum values in the comparative clam determination device 105, respectively, and similarly, standard data BM and standard data
BN is the maximum value and minimum value in the comparison/judgment circuit 106, respectively. At this time, the reference data AM,
BM corresponds to the first value or third value recited in the claims, and the reference data AN, BN correspond to the second value or fourth value recited in the claims. Up until now, in order to make the explanation easier, the solid-state image sensor 2
Although one pixel column is assumed as shown in FIG. 1a, the image signal detection and correction apparatus of the present invention can of course be used for other pixel columns. As the pixel array of the solid-state image sensing device, other arrangement arrays such as those shown in FIGS. 11a and 11b are also conceivable. In this case as well, the pixel signal block A _i
B _i C _i 111 and A _i B _i C _i D _i 112 can be considered as one set and signal processing can be performed sequentially. Also, at this time, the image signal does not necessarily have to be the pixel signal or color difference signal used in the explanation; for example, R,
It is also possible to remove false signals after checking the G and B signals. FIG. 12a shows a specific example of the pixel signal arrangement of FIG. 11b. W _i , G _i , Cy _i , and YYe _i represent the i-th white, green, cyan, and yellow signal levels, respectively. An R signal and a B signal are extracted from the i-th pixel signal block 121 by the following calculation. R signal...W _i -Cy _i , Ye _i -G _i B signal...Cy _i -G _i , W _i -Ye _i At this time, luminance signal change 1 as shown in Fig. 12b
22, the B signal becomes a false signal. Since the pixel signals W _i and Cy _i and the pixel signals G _i and Ye _i have the same luminance level, there is no problem in the calculations W _i −Cy _i and Ye _i −G _i for creating the R signal. but,
The B signal obtained by performing calculations on Gy _i and G _i and W _i and Ye _i , which have different brightness states, becomes a false signal. Therefore, using Cy _i −G _i-1 instead of Cy _i −G _i ,
Furthermore, false signals can be removed by using W _i+1 −Ye _i instead of W _i −Ye _i . That is, the change 122 in the luminance signal is compared with the reference data, and if correction is necessary, G _i and W _i are replaced with G _i-1 and W _i+1, respectively, in the signal processing to obtain the B signal. Bye. Note that in the embodiment of the present invention, all false signals are treated as appearing at the boundaries in the horizontal direction; however,
Of course, there are also vertical boundary edges in various objects. At this time, if the above-described image signal detection and correction apparatus of the present invention is applied in the vertical direction of the screen using a plurality of delay lines of about 1H, false signals occurring in the vertical direction can be removed. Effects of the Invention As described above, the image signal detection and correction device of the present invention detects the signal level change between the h-th image signal and the h+1th and subsequent image signals generated point-sequentially from the solid-state image sensor, and this level change is When the value is larger than a certain set reference data, the image signal generated point-sequentially from the solid-state image sensor is replaced with another image signal, thereby making it possible to remove false signals occurring at the boundaries of the screen. Therefore, color errors around the boundary edges are eliminated, and the image quality can be significantly improved, which is very effective.

[Brief explanation of the drawing]

FIG. 1 is an explanatory diagram showing a conventional method for extracting color difference signals, and FIGS. 2a and 2b are overall block diagrams showing an embodiment of the image signal detection and correction device of the present invention, respectively.
3a and 3b are explanatory diagrams showing the operating principle of the overall block diagram of FIGS. 2a and 2b, FIG. 4 is a specific configuration diagram of a color signal processing device, and FIG. 5 is an illustration of an image signal processing device. FIGS. 6a and 6b are specific configuration diagrams of the image signal level change detection device, FIGS. 7a and 7b are configuration diagrams of the comparison and determination device, and FIG. 8 is a diagram of the image signal replacement device. One specific configuration diagram, FIG. 9 is an explanatory diagram showing the operation when pixel signals are used as image signals in the image signal detection and correction device of the present invention, FIG. 10 is a block configuration diagram thereof, and FIGS. 11a and b 12 is a schematic diagram showing a pixel row of a solid-state image sensor, and FIG. 12 is a schematic diagram showing a specific example of the pixel row in FIG. 11b. 21... Solid-state image sensor, 22... Color signal processing device, 23... Image signal processing device, 24... Image signal level change detection device, 25... Comparison and determination device,
26... Image signal replacement device, 27, 73, 74,
78, 79, 109, 110...control signal.

Claims (1)

[Claims] 1. Level change detection means for obtaining level difference data between the h-th image signal and the (h+1)-th image signal (h is a natural number) generated point-sequentially from the solid-state image sensor; The first step to determine whether the value is greater than or equal to
The level difference data is compared with the first and second reference data, and the level difference data is compared with the first and second reference data. is provided with a comparison determination means for determining whether the level difference data is greater than or equal to the first reference data or less than the second reference data, and when the level difference data is greater than or equal to the first reference data, the h-th image created point-sequentially from the solid-state image sensor When the signal is replaced with the (h+1)th image signal or later, and the level difference data is less than or equal to the second reference data, the hth image signal generated dot-sequentially from the solid-state image sensor is replaced with (h+1)th image signal.
-1) An image signal detection and correction device provided with an image signal replacement means for replacing the image signal with the previous image signal. 2. Claim 1, wherein the image signal generated point-sequentially from the solid-state image sensor is a pixel signal directly extracted from the solid-state image sensor.
The image signal detection and correction device described in . 3. The image signal produced point-sequentially from the solid-state image sensor is a signal obtained by passing a pixel signal directly extracted from the solid-state image sensor through a low-pass filter. Image signal detection correction device. 4. The image signal detection and correction device according to claim 1, wherein the image signal generated dot-sequentially from the solid-state image sensor is a color difference signal. 5. The image signal detection and correction device according to claim 1, wherein the image signals generated dot-sequentially from the solid-state image sensor are a red R signal, a green G signal, and a blue B signal. 6. A level change detection means is provided to obtain a ratio value between the h-th image signal and the (h+1)-th image signal (h is a natural number) generated point-sequentially from the solid-state image sensor, and determines a value equal to or higher than the third value. and fourth reference data for determining the ratio value and the third and fourth values.
The value of the ratio is compared with the reference data of the third
is provided with a comparison/judgment means for determining whether the ratio is greater than or equal to the fourth reference data, and when the value of the ratio is greater than or equal to the third reference data, the h-th image produced point-sequentially from the solid-state image sensor is provided. Replace the signal with the (h+1)th or later image signal, and when the value of the ratio is less than or equal to fourth reference data, replace the signal with the (h-1)th or earlier image signal generated point-sequentially from the solid-state image sensor. An image signal detection and correction device provided with an image signal replacement means. 7. Claim 6, characterized in that the image signal generated point-sequentially from the solid-state image sensor is a pixel signal directly extracted from the solid-state image sensor.
The image signal detection and correction device described in . 8. The image signal produced point-sequentially from the solid-state image sensor is a signal obtained by passing a pixel signal directly extracted from the solid-state image sensor through a low-pass filter. Image signal detection correction device. 9. The image signal detection and correction device according to claim 6, wherein the image signal generated dot-sequentially from the solid-state image sensor is a color difference signal. 10. The image signal detection and correction apparatus according to claim 6, wherein the image signals generated dot-sequentially from the solid-state image sensor are a red R signal, a green G signal, and a blue B signal.