JPH0523545B2 - - Google Patents

Description

[Detailed description of the invention]

INDUSTRIAL APPLICATION FIELD The present invention relates to an apparatus for processing a video signal captured by a television camera (hereinafter referred to as a TV camera). BACKGROUND ART Conventionally, TV cameras have been put into practical use as inputs for broadcasting or home VTRs. In a TV camera, the optical image of the subject is photoelectrically converted using an image pickup tube or image pickup plate, the output is passed through a matrix circuit to obtain luminance signals and color signals, and these signals are converted into composite video signals such as NTSC and PAL. Send to outside. Problems to be Solved by the Invention This TV camera is operated by a cameraman when used for broadcasting, and by a general consumer when used for home use. There is not much of a problem when operating a TV camera fixed on a tripod, etc., but general consumers who are not familiar with the handling of TV cameras may try to hold the camera still while taking images by hand-held shooting. In this case, "camera shake" occurs, and when the output signal of the TV camera at this time is displayed on a monitor TV,
The image vibrates vertically and horizontally, resulting in an extremely unsightly image. In addition, even in the case of signals used for broadcasting operated by experts, when runners are imaged by a TV camera mounted on a relay vehicle in the case of a marathon broadcast, etc., the vibrations of the vehicle traveling are transmitted to the TV camera, causing a large difference in the video image. Vertical movement occurs, resulting in an unsightly image. Furthermore, with portable VTRs (ENG) used for news gathering, it is practically impossible for a cameraman to take images while running due to the same problem of camera shake. The present invention solves these problems and provides a video signal processing device that performs signal processing such that even when an image is captured by a TV camera with camera shake, the influence of camera shake does not occur on the image. Means for Solving the Problems The present invention provides a first television scanning system in which the number of scanning lines is N, the frame period T _F , and _the horizontal scanning period _TH . An imaging means operated by a second scanning system with a frame period T _F horizontal scanning period T _H /K _V and a scanning range expanded by K _H times in the horizontal direction and K _V times in the vertical direction; means for A/D converting the output of the image pickup means; a screen memory for storing a plurality of screens of the output of the imaging means digitized by the A/D conversion means;
It consists of a motion detection means that detects the movement of the entire screen by comparing the contents of the image memory with the screen being imaged, and when reading out part of the contents of the screen memory, the output of the motion detection means is used to detect the movement of the entire screen. This controls the reading location. Effects According to the present invention, with the above-mentioned configuration, even when an image is captured by an imaging means that is subjected to motion that is harmful to the image, it is possible to obtain an image free of the influence of the motion. That is, the imaging means is being scanned by the second television. This is K _V times the vertical direction and K _H times the horizontal direction compared to the first television scanning system that should be obtained as the final output.
It scans twice as wide an area. The image from the second scanning system, ie, wide range scanning, is A/D converted and stored in screen memory. This screen memory can store multiple screens. If the imaging means moves during imaging, the image being captured will move relative to the image stored in the screen memory. This motion is detected by motion detection means. 1st
The television scanning system scans a part of the screen of the second television scanning system, but when reading from the screen memory, the reading range is moved in accordance with the movement detected by the movement detection means. In this case, the image that was moving in the second television scanning system appears to be stationary in the first television scanning system. Therefore, the first television scanning system is
If the television signal processing system of the present invention is applied as a standard system such as NTSC or PAL,
Even if a subject is imaged by a television camera that is subject to harmful motion, it will appear as if it were a stationary NTSC or
You can obtain images that look like they were taken with a PAL television camera. Embodiment Before describing a specific embodiment, an outline of the operating principle will be explained with reference to FIG. 2. Consider a case where an image pickup tube is used as the image pickup device. In FIG. 2, 1 indicates the target of the image pickup tube. The range indicated by 2 is the scanning range required when transmitting a normal television signal. Therefore, a normal NTSC television camera scans this range at a horizontal scanning frequency of 15.734 KHz, a vertical scanning frequency of 59.9 Hz, and 525 scanning lines (excluding blanking periods, to be exact). In contrast, the present invention attempts to scan the range indicated by 3 on the target 1, and after signal processing, transmit the range indicated by 2. Further, it is assumed that the position of range 2 within range 3 can be freely moved as indicated by 2'. The direction and amount of movement are determined so as to follow unnecessary image movement within range 3 caused by vibration of the camera during imaging. As a result, even if the image moves within range 3, range 2
The transmitted image is obtained assuming that the image is at a fixed position within the image. FIG. 1 shows an embodiment of the present invention. 4 is an image pickup tube;
5 is a frame memory capable of storing three screens, 6 is a synchronization signal generation circuit, and 7 is a motion detection circuit. The image pickup tube 4 is operated by a synchronizing signal S _y 2 (second scanning system) generated by a synchronizing signal generating circuit 6, and scans a scanning range 3. This image pickup output is written into the frame memory 5 as it is if it is an analog memory, or digitized and written into the frame memory 5 if it is a digital memory. The operating timing of the frame memory at this time is called a mode. FIG. 3 shows an explanatory diagram of the operation timing of the frame memory. In this figure, M to M+3 indicate frame numbers. When imaging the Mth frame, this frame is written into the frame memory 5 at the same time. Reading from the frame memory 5 is performed at the same time as writing, and this involves reading out a signal from one frame before at exactly the same position as the position being imaged by the image pickup tube 4 at the same mode timing. The signal written in the frame memory, that is, the M frame signal, and the signal read out from the frame memory, that is, (M-
1) Frame signals are input to the motion detection circuit 7, and motion vectors between these frames are determined. The motion detection circuit has a buffer memory corresponding to the capacity of a small part of the screen, and relatively moves the position of the blocks in one frame among the small blocks in the two frames to find the position and direction where the correlation is greatest. The above motion vector is detected by this. For such small blocks, motion vectors may be calculated at several positions within one frame, and the motion of the entire screen may be detected using majority logic. As described above, at the timing of imaging frame No. (M), it is determined "in which direction and by how much distance" frame M has moved compared to frame (M-1). The imaging frame is determined according to this result.
At timing No. (M+2), the signal of frame M is read out only in scanning range 2 in FIG. The operating timing of the frame memory at this time is called a mode. This operates based on the synchronization signal S _y1 (first scanning system). The operation timing and mode of the frame memory will be explained below. The table below shows the specifications of the mode when obtaining an NTSC signal as the final output signal.

[Table] Here, it is clear that the larger K _V and K _H are, the greater the movement between two adjacent frames can be compensated for. However, as K _V and K _H are increased, the highest frequency of the video signal increases accordingly. The target area of the image pickup tube must be increased. If the size of the target is not changed, the resolution will deteriorate. Problems such as these will become even bigger. For, the highest frequency in scanning system 1 is
If _nax1 , the highest frequency in scanning system 2 is
_nax2 = K _V × K _H × _nax1 . Also, since the area of the target is also multiplied by K _V × K _H ,
Practically speaking, it is considered appropriate that K=K _V =K _H = about 1.2 to 1.6. Differences in operation depending on the frame memory mode when K=1.5 will be explained below. FIG. 4a shows the vertical operation timing. The shaded portion in the timing shown in the mode corresponds to scanning range 2 in the mode. This part is read out in mode and expanded on the time axis. Further, the shaded portion of the mode indicates that the mode moves from the lower limit to the upper limit of the scanning range 3 between the M frame and the (M+2) frame. FIG. 4b shows the horizontal movement timing. SyNC is the horizontal synchronization signal period. Although not shown in the figure, there is a time difference of two frames between writing and reading in this case, as can be seen from a. This figure shows that the Nth scanning line of the mode corresponds to the (ND)th scanning line of the mode. This is an example where the first scan line of scan range 2 starts from the (D+1)th scan line of scan range 3. In this way, since the frame memory is read in a narrower range both vertically and horizontally than the image captured by the second scanning system, it is necessary to perform the above-mentioned time axis expansion process in order to obtain a normal image using the first scanning system. There is. As is clear from the above description, according to the present invention, by changing the reading position of the frame memory in accordance with unnecessary movement of the camera, it is possible to obtain an image as if it were taken from a stationary camera. On the other hand, when the camera is intentionally moved, that is, when camera operations such as panning and tilting are being performed, inconveniences occur if the signal processing of the present invention is operated. In order to prevent this situation, the camera is equipped with a manual switch that can distinguish between panning, tilting, and stilling, and when performing a certain operation, the appropriate switch can be turned on. It is sufficient to provide a control signal that compensates only for horizontal movement and for vertical and horizontal movement when the vehicle is stationary. It is also necessary to provide a switch to disable this signal processing during complicated camera operations such as zooming. Furthermore, in order to fully support cases where complex camera operations are performed such as for broadcasting, a screen that supports scanning range 3 should be designed for this signal processing.
It is conceivable to record the signal once on a VTR, perform the main signal processing on the reproduced signal, and then edit and re-record it on a standard VTR. The advantage of having such a process is that if the accuracy of the movement detected by the motion detection circuit is not good, the correct movement can be given according to the editor's instructions, and it can fully handle complex camera work. It is possible. Further, in all the above explanations, movement between frames has been considered as a problem, but in order to deal with more intense movement, it is conceivable to perform similar processing between fields. Effects of the Invention As detailed above, according to the present invention, images obtained by hand-held imaging by general consumers, or images captured by a TV camera installed in a car with strong vibrations, etc. It is possible to remove unnecessary blur from an image containing blur unrelated to the original camera work and restore it to a normal image. As a result, not only can you obtain much better images than before in cases such as those mentioned above, but you can also conduct camera interviews that were previously impossible for practical purposes, such as holding a TV camera and taking images while driving. .

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing one embodiment of the present invention;
FIG. 2 is a diagram showing the relationship between the imaging screen and the sending screen, FIG. 3 is an operation timing diagram in one embodiment of the present invention, and FIG. 4 is a more detailed timing diagram in one embodiment of the present invention. 4... Image pickup tube, 5... Frame memory, 6...
Synchronization signal generation circuit, 7...Motion detection circuit.

Claims (1)

[Scope of Claims] 1. An imaging means, a memory means for storing screen information that is an output of the imaging means, and a motion detection means for detecting movement of an image over the entire screen between temporally consecutive screens. , reads screen information from a reading position in a range equivalent to 1/K _U times (K _U > 1) in the vertical direction and 1/K _H times (K _H > 1) in the horizontal direction of the contents of the above screen memory. and moving the readout position according to the output of the motion detection means so as to cancel out the detected motion, thereby removing vibrations in the image due to blurring of the image pickup means. television signal processing equipment. 2. The television signal processing device according to claim 1, wherein 1.2<K _U <1.6 and 1.2<K _H <1.6. 3. The television signal processing device according to claim 1, wherein the screen memory is a frame memory. 4. The television signal processing device according to claim 1, wherein the screen memory is a field memory. 5. The television signal processing device according to claim 1, wherein the imaging means includes an imaging tube. 6. The television signal processing device according to claim 1, wherein the imaging means includes a solid-state imaging plate. 7 If the first scanning means is NTSC or PAL or
The television signal processing device according to claim 1, characterized in that it conforms to the SECAM system. 8. The television signal processing device according to claim 1, further comprising means for time-axis expanding the read signal. 9. The television signal processing apparatus according to claim 1, further comprising means for setting a mode for restricting the operation of the control means. 10. The television signal processing apparatus according to claim 1, further comprising means for setting a mode in which the operation of the control means is stopped. 11. The television signal processing apparatus according to claim 1, wherein a motion vector is detected using information of a plurality of small blocks between two temporally consecutive screens.