JP3861717B2 - Network camera and image distribution processing device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a network camera and an image distribution processing device for distributing image data to a client via a network such as the Internet.
[0002]
[Prior art]
In recent years, image distribution systems that distribute image data to clients via a network such as the Internet have begun to spread. This type of image distribution system includes a video-on-demand system that transmits image data while reading out image data from a pre-stored image file, and a real-time broadcast-type system that transmits images from a video camera or the like after compression encoding. Divided into systems.
[0003]
A real-time broadcast type image distribution system includes a video camera that inputs an image, an image server that distributes an image given from the video camera via a network such as the Internet, and a plurality of receiving clients that receive an image distributed from the image server It is configured as a client / server system consisting of Using a network camera having a network connection function as a camera server, an image can be distributed from the camera server to a plurality of receiving clients connected to a network such as the Internet.
[0004]
The network camera used as the camera server has a configuration as shown in FIG.
[0005]
That is, the network camera 100 shown in FIG. 8 includes a camera block 110 for imaging, a signal processing unit 120 to which imaging output from the camera block 110 is supplied, a communication interface unit 130 for communication by connecting to an external network, It consists of a system control unit 140 for controlling these.
[0006]
The camera block 110 includes an interline CCD image sensor 111, a timing generator 112 for operating the interline CCD image sensor 111, a front end unit 113 to which an imaging output from the interline CCD image sensor 111 is supplied, A digital signal processing unit 114 to which the output of the front end unit 113 is supplied, a camera for controlling the operation of the digital signal processing unit 114, white balance, focusing, etc. based on information from the digital signal processing unit 114 It is composed of a camera control CPU 115 and the like for performing control.
[0007]
In the camera block 110, the interline CCD image sensor 111 performs photoelectric conversion in a field cycle, and alternately outputs an interlaced output, that is, an odd field imaging output and an even field imaging output constituting one frame. The analog imaging signal is sequentially output as the imaging output of each field. The front end unit 113 includes an A / D converter that samples and holds an analog imaging signal from the interline CCD image sensor 111 and performs analog / digital conversion. The digital signal processing unit 114 performs digital signal processing on the digital imaging signal from the front end unit 113 and outputs an interlaced YUV (4: 1: 1) digital image signal.
[0008]
The signal processing unit 120 is supplied with an interlaced YUV (4: 1: 1) digital image signal from the digital signal processing unit 114 of the camera block 110, and a light to which an output of the format conversion unit 117 is supplied. FIFO memory 118, frame memory write controller 119 to which the output of this write FIFO memory 118 is supplied, frame memory interface unit 121 to which the output of this frame memory write controller 119 is supplied, and a frame connected to this frame memory interface unit 121 Memory 131, frame image frame memory read controller 122 to which the output of the frame memory interface unit 121 is supplied, and field image frame memory read controller 1 3. A read FIFO memory 125 to which an output of the frame image frame memory read controller 122 or the field image frame memory read controller 123 is selectively supplied via an image selector 124, and an output of the read FIFO memory 125 is supplied. JPEG / I / F unit 126, JPEG compression processing unit 127 and line buffer memory 132 connected to the JPEG / I / F unit 126, and CPU / I / F to which the output of the JPEG compression processing unit 127 is supplied Part 128, and an image selection bit register 129 connected to the CPU / I / F part 128.
[0009]
In the signal processor 120, the format converter 117 converts the interlaced YUV (4: 1: 1) digital image signal from the digital signal processor 114 into a YCbCr (4: 2: 2) signal for JPEG compression. Convert.
[0010]
The write FIFO memory 118 absorbs the difference between the signal speed from the camera block 110 and the writing speed to the frame memory 131. The frame memory write controller 119 converts the YCbCr (4: 2: 2) signal supplied from the format converter 117 via the write FIFO memory 118 from an interlace signal to a non-interlace signal. Control is performed to write to the frame memory 131 via the frame memory interface unit 121.
[0011]
The frame memory interface unit 121 controls writing, reading, and refreshing to the frame memory 131 in a time division manner.
[0012]
The frame image frame memory read controller 122 controls the frame memory 131 to read the first field image and the second field image as one frame image via the frame memory interface unit 121.
[0013]
The field image frame memory read controller 123 controls to read only the first field image or the second field image from the frame memory 131 via the frame memory interface unit 121.
[0014]
The image selector 124 receives the frame based on the Sel signal based on the image selection bit supplied from the system control CPU 141 of the system control unit 140 to the image selection bit register 129 via the CPU / I / F unit 128. The frame image YCbCr (4: 2: 2) signal read from the frame memory 131 by the image frame memory read controller 122 or the field image read from the frame memory 131 by the field image frame memory read controller 123. Select the YCbCr (4: 2: 2) signal.
[0015]
The read FIFO memory 125 absorbs the difference between the YCbCr (4: 2: 2) signal speed of the frame image or field image read from the frame memory 131 and the JPEG compression speed. The JPEG I / F unit 126 temporarily stores the YCbCr (4: 2: 2) signal of the frame image or field image supplied via the image selector 124 in the line buffer memory 132, and performs raster block conversion. The data is input to the JPEG compression processing unit 127 and the operation of the JPEG compression processing unit 127 is controlled. The JPEG compression processing unit 127 performs compression processing based on JPEG on the raster block converted YCbCr (4: 2: 2) signal. The output of the JPEG compression processing unit 127, that is, JPEG compressed image data is written to the system RAM 142 of the system control unit 140 via the CPU / I / F unit 128.
[0016]
The JPEG compressed image data is written into the system RAM 142 using, for example, the DMA function of the system control CPU or via the system control CPU 141.
[0017]
The communication interface unit 130 includes an Ethernet controller (MAC) 135 for a network interface, a physical layer chip 136, a pulse transformer 137 for impedance matching, an RJ-45 connector 138, and the like.
[0018]
The system control unit 140 includes a system control CPU 141, a buffer and working area system RAM 142, and a program system ROM 143 connected via a bus. The system control CPU 141 controls the system and camera control CPU 115. Communicate.
[0019]
The network camera 100 is equipped with a protocol stack such as TCP / IP or HTTP as software, and can remotely view images and control the camera with a viewer such as Internet Explorer (IE). Yes. An image size can be selected in the camera control menu. For example, the user can select an image of 640 × 480 (frame image) or 640 × 240 (field image).
[0020]
[Problems to be solved by the invention]
By the way, in the conventional network camera 100 configured as described above, when an image size (frame image) of 640 × 480 is selected, when a stationary object that does not move is a subject, it is more vertical than 640 × 240 (field image). However, for moving subjects, the first field image and the second field image are shifted in the horizontal direction because the accumulation timing of the interline CCD image sensor 111 is different. As a result, there is a problem that the image quality deteriorates. In order to avoid this, when an image size of 640 × 240 (field image), that is, a field image is selected, a vertical line is simply doubled and displayed on the viewer side. However, the image is not horizontally shifted by the interline CCD image sensor 111 as in the case of a stationary object, but the image quality deteriorates because the vertical resolution is lower than that of 640 × 480 for a stationary object. It was.
[0021]
Further, since the image signal speed from the camera block 110 (writing speed to the frame memory 131) and the processing system speed (reading speed from the frame memory 131) are different, the next frame image writing is being performed while one frame image is being read. In the case of an object that has overtaken and moved, overtaking operation noise has occurred.
[0022]
In view of the above-described conventional problems, an object of the present invention is to obtain a high-quality image output when distributing image data supplied from an interlaced image data supply device to a client via a network. There is to be able to do it. Another object of the present invention is to eliminate overtaking scanning noise and obtain a high-quality image output. Furthermore, an object of the present invention is to make it possible to adaptively perform switching between a field image and a frame image by following setting information such as a gain value of an imaging means.
[0023]
[Means for Solving the Problems]
The network camera according to the present invention includes an interlaced output imaging unit, a frame image generating unit and a field image generating unit that generate non-interlaced frame image data and field image data from image data supplied from the imaging unit, The moving object detection means for detecting the moving object of the subject based on the image data supplied from the imaging means, and the image data supplied from the imaging means based on the detection output from the moving object detection means is still image data. Whether the image data supplied from the image pickup means is image data of a still image, the frame image data generated by the frame image creation means is selected, and the image pickup means When the image data supplied from the Obtained by the image data selection means for selecting the field image data generated by the yield image creation means, the compression processing means for converting the image data selected by the image data selection means into compressed image data, and the compression processing means Network interface means for sending compressed image data to the network, a frame memory having a storage area for a plurality of frames, and at the start of writing image data to the frame memory, select an area where image data is not read, A memory control means for selecting an area where writing of image data has been completed for one or more fields at the start of reading image data from the frame memory and writing / reading image data to / from the frame memory; The frame image creation means generates frame image data by reading out the image data supplied from the imaging means as non-interlaced frame image data via the frame memory, and the network interface means is connected to the Internet. The compressed image data is distributed in real time based on the selected image size to a plurality of clients connected to the Internet. It is characterized by that.
[0024]
The present invention also provides an image distribution processing device for distributing image data supplied from an interlaced output image data supply device to a client via a network, wherein the image data supplied from the image data supply device is non-interlaced. Frame image creation means and field image creation means for generating frame image data and field image data, moving object detection means for detecting a moving object of a subject based on image data supplied from the imaging means, and the moving object detection means Based on the detection output, it is determined whether the image data supplied from the imaging unit is still image data or moving image data, and the image data supplied from the imaging unit is still image data. Is the frame image data generated by the frame image generating means. And image data selection means for selecting field image data generated by the field image creation means when the image data supplied from the imaging means is moving image data, and the image data selection means. Compression processing means for converting the selected image data into compressed image data, network interface means for sending the compressed image data obtained by the compression processing means to the network, a frame memory having a storage area for a plurality of frames, and the above When starting to write image data to the frame memory, select the area where image data has not been read. When starting to read image data from the frame memory, select the area where writing of the image data has been completed for one or more fields. To the above frame memory A memory control means for writing / reading image data, and the frame image creation means reads out the image data supplied from the imaging means as non-interlaced frame image data via the frame memory, Frame image data is generated, and the network interface means is connected to the Internet and distributes the compressed image data in real time based on the selected image size to a plurality of clients connected to the Internet. To do.
[0025]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
[0026]
The network camera according to the present invention is configured, for example, as shown in FIG.
[0027]
That is, the network camera 1 shown in FIG. 1 includes a camera block 10 for imaging, a signal processing unit 20 to which imaging output from the camera block 10 is supplied, a communication interface unit 30 for communication by connecting to an external network, It consists of a system control unit 40 for controlling these.
[0028]
The camera block 10 includes an interline CCD image sensor 11, a timing generator 12 for operating the interline CCD image sensor 11, a front end unit 13 to which an imaging output from the interline CCD image sensor 11 is supplied, A digital signal processing unit 14 to which the output of the front end unit 13 is supplied, a camera for controlling the operation of the digital signal processing unit 14 and white balance, focusing, etc. based on information from the digital signal processing unit 14 The camera control CPU 15 is used for control.
[0029]
In the camera block 10, the interline CCD image sensor 11 performs photoelectric conversion in a field cycle, and alternately outputs an interlaced output, that is, an odd field imaging output and an even field imaging output constituting one frame. The analog imaging signal is sequentially output as the imaging output of each field. The front end unit 13 includes an A / D converter that samples and holds an analog imaging signal from the interline CCD image sensor 11 and performs analog / digital conversion. The digital signal processing unit 14 performs digital signal processing on the digital imaging signal from the front end unit 13 and outputs an interlaced YUV (4: 1: 1) digital image signal.
[0030]
The signal processing unit 20 includes a frame synchronization signal generation unit 16 that generates a frame synchronization signal XVD and a field signal FLD based on the vertical synchronization signal VD supplied from the digital signal processing unit 14 of the camera block 10, and the camera block 10. An interlaced YUV (4: 1: 1) digital image signal, a vertical synchronizing signal VD, a horizontal synchronizing signal HD, a field signal FLD, and a camera synchronizing clock CCLK are supplied from the digital signal processing unit 14 and from the frame synchronizing signal generating unit 16. The format conversion unit 17 to which the frame synchronization signal XVD is supplied, the write FIFO memory 18 to which the output of the format conversion unit 17 is supplied, the frame memory write controller 19 to which the output of the write FIFO memory 18 is supplied, and the frame memory write A frame memory interface unit 21 to which an output of the controller 19 is supplied, a frame memory read controller 31 and a frame memory 34 connected to the frame memory interface unit 21, and a frame to which an output of the frame memory interface unit 21 is supplied The outputs of the image frame memory read controller 22 and the field image frame memory read controller 23, the frame image frame memory read controller 22, or the field image frame memory read controller 23 are selectively supplied via an image selector 24. Read FIFO memory 25, the JPEG I / F unit 26 to which the output of the read FIFO memory 25 is supplied, and a JPEG compression process connected to the JPEG I / F unit 26 Unit 27 and line buffer memory 35, CPU / I / F unit 28 to which the output of the JPEG compression processing unit 27 is supplied, image selection bit register 29 connected to the CPU / I / F unit 28, and moving object detection A moving object detection FIFO memory 32 to which the output of the frame memory read controller 31 is supplied, an output of the moving object detection FIFO memory 32, a moving object detector 33 to which the output of the write FIFO memory 18 is supplied, and the like.
[0031]
In the signal processing unit 20, the frame synchronization signal generation unit 16 generates a frame synchronization signal XVD and a field signal FLD generated based on the vertical synchronization signal VD of the field period supplied from the digital signal processing unit 14 of the camera block 10. The data is supplied to the format conversion unit 17 and the frame memory write controller 19 and is also supplied to the system control CPU 41 of the system control unit 40.
[0032]
The format converter 17 converts the interlaced YUV (4: 1: 1) digital image signal from the digital signal processor 14 into a YCbCr (4: 2: 2) signal for JPEG compression.
[0033]
The write FIFO memory 18 absorbs the difference between the signal speed from the camera block 10 and the writing speed to the frame memory 34.
[0034]
The frame memory write controller 19 converts the YCbCr (4: 2: 2) signal supplied from the format conversion unit 17 via the write FIFO memory 18 from an interlace signal to a non-interlace signal. Control is performed to write to the frame memory 34 via the frame memory interface unit 21.
[0035]
The frame memory interface unit 21 controls writing, reading, and refreshing to the frame memory 34 in a time-sharing manner. An example of memory control timing by the frame memory interface unit 21 is shown in FIG.
[0036]
The frame image frame memory read controller 22 controls to read out the first field image data and the second field image data as one frame image data from the frame memory 34 via the frame memory interface unit 21. .
[0037]
The field image frame memory read controller 23 reads only the image data of the first field or the image data of the second field from the frame memory 34 via the frame memory interface unit 21, and Create and interpolate average data of line data.
[0038]
In the network camera 1, the frame memory 34 provides an image area of at least two frames.
[0039]
Then, the motion detection frame memory controller 31 synchronizes with the image data supplied from the camera block 10 to the motion detector 33 via the write FIFO memory 18 for motion detection, from the frame memory 34 to the frame memory. Image data one frame before is read out via the interface unit 21.
[0040]
The motion detection FIFO memory 32 is configured to display the speed of image data supplied to the motion detector 33 via the write FIFO memory 18 and the image one frame before read from the frame memory 34 by the motion detection frame memory controller 31. Absorb data speed differences.
[0041]
The motion detector 33 compares the image data supplied from the camera block 10 via the write FIFO memory 18 with the image data of the previous frame read from the frame memory 34 by the motion detection frame memory controller 31. Thus, a moving object detection output is obtained.
[0042]
The moving body detector 33 is configured as shown in FIG. 3, for example.
[0043]
That is, the moving body detector 33 shown in FIG. 3 includes a difference threshold value register 331 in which a difference threshold value TH is set from the system control CPU 41 of the system control unit 40 via the CPU / I / F unit 28, and the camera block 10. The absolute difference between the image (luminance) data Yc supplied via the write FIFO memory 18 and the image (luminance) data Yo one frame before read from the frame memory 34 by the frame memory controller 31 for motion detection. A circuit 332, a comparator 333 that compares the difference threshold TH set in the difference threshold register 331 with the difference absolute value obtained by the difference absolute value circuit 332, and the number of pixels that perform a counting operation according to the comparator 333 A counter 334 and a pixel number register to which the count value of the pixel number counter 334 is given Consisting of 35.
[0044]
In this motion detector 33, image (luminance) data Yc supplied from the camera block 10 via the write FIFO memory 18 and an image one frame before read from the frame memory 34 by the motion detection frame memory controller 31 are displayed. The absolute value ABSΔ of the difference from the (luminance) data Yo is obtained by the difference absolute value circuit 332, and the difference absolute value ABSΔ obtained by the difference absolute value circuit 332 is set to the difference threshold value TH set in the difference threshold value register 331 and the above value. By comparing with the comparator 233, a pixel that gives a difference absolute value ABSΔ larger than the difference threshold TH is detected as a pixel that has moved in the image, and a pixel number counter that is reset in the frame period by the frame synchronization signal XVD Count at 334. The count value of the pixel number counter 334, that is, the number of pixels giving the difference absolute value ABSΔ larger than the difference threshold TH is copied to the pixel number register 325 before being reset by the frame synchronization signal XVD.
[0045]
The pixel number count value copied to the pixel number register 325 is read by the system control CPU 41 of the system control unit 40 via the CPU / I / F unit 28.
[0046]
The system control CPU 41 of the system control unit 40 determines whether the image output from the camera block 10 is a moving image but a still image based on the pixel count value copied to the pixel count register 325. Can be determined.
[0047]
The image selector 24 receives the selection signal based on the image selection bit supplied from the system control CPU 41 of the system control unit 40 to the image selection bit register 29 via the CPU / I / F unit 28. The frame image YCbCr (4: 2: 2) signal read from the frame memory 34 by the frame image frame memory read controller 22 or the field image read from the frame memory 34 by the field image frame memory read controller 23. YCbCr (4: 2: 2) signal is selected.
[0048]
The read FIFO memory 25 absorbs the difference between the speed of the YCbCr (4: 2: 2) signal of the frame image or field image read from the frame memory 34 and the JPEG compression speed. The above JPEG. The I / F unit 26 temporarily stores the YCbCr (4: 2: 2) signal of the frame image or field image supplied via the image selector 24 in the line buffer memory 35, and performs raster block conversion. The input to the JPEG compression processing unit 27 and the operation of the JPEG compression processing unit 27 are controlled. The JPEG compression processing unit 27 performs compression processing conforming to JPEG on the raster block-converted YCbCr (4: 2: 2) signal. The output of the JPEG compression processing unit 27, that is, JPEG compressed image data is written into the system RAM 42 of the system control unit 40 via the CPU / I / F unit 28.
[0049]
The JPEG compressed image data is written into the system RAM 42 using, for example, the DMA function of the system control CPU 41 or via the system control CPU 41.
[0050]
The communication interface unit 30 includes an Ethernet controller (MAC) 35 for a network interface, a physical layer chip 36, a pulse transformer 37 for impedance matching, an RJ-45 connector 38, and the like.
[0051]
The system control unit 40 includes a system control CPU 41, a buffer and working area system RAM 42, and a program system ROM 43 connected via a bus. The system control CPU 41 and the system control and camera control CPU 15 Communicate.
[0052]
The network camera 1 is equipped with a protocol stack such as TCP / IP, HTTP, etc. as software, and can remotely view images and control the camera with a viewer such as Internet Explorer (IE). Yes. An image size can be selected in the camera control menu. For example, the user can select an image of 640 × 480 (field image) or 640 × 240 (field image).
[0053]
In the network camera 1 having such a configuration, the frame image / field image automatic switching operation is performed according to the procedure shown in the flowchart of FIG.
[0054]
In this network camera 1, when the image size is selected as 640 × 480 (Auto) in the menu, the frame counter value and the field counter value are set to 0 as initialization (step S1).
[0055]
Next, the system control CPU 41 performs RS232C communication with the camera control CPU 15, obtains the current gain information of the camera block 10, and saves the value in the previous Gain register. Then, the difference threshold TH corresponding to the Gain is written in the difference threshold of the moving object detector 33 (step S2). For example, when the Gain setting of the camera block 10 is set to a high value of +18 db, the difference threshold TH is also set to a high value, and when the Gain setting value is low to 0 db, the difference threshold TH is also set to a low value. When the Gain setting value is high, the noise level is also high. Therefore, the gain information is dynamically fed back to the moving body detector 33 in order to remove the influence of the noise of the moving body detection result and improve the accuracy of moving body detection. Yes.
[0056]
Next, the system control CPU 41 communicates with the camera control CPU 15, obtains the current information of the camera block 10, and saves it in the current Gain register (step S3). Then, the previous Gain and the current Gain register value are compared (step S4). If they are different, the current Gain value is written in the previous Gain register (step S5), and the difference threshold value of the moving object detector 33 corresponding to the current Gain is stored in the register. Write (step S6). If the previous Gain and the current Gain value are the same, nothing is done. Then, it waits twice for a frame cycle interruption by the frame synchronization signal XVD (step S7), and reads the pixel number counter value from the pixel number register 335 from the moving object detector 33 (step S8). Then, the pixel number counter value is compared with the moving object detection sensitivity value (this value can be set from the viewer menu) (step S9). If the count value is smaller than the set moving object detection sensitivity value, the count value is set. If it is smaller than the motion detection sensitivity value, it is determined whether or not the field counter value is 0 (step S10). If the field counter value is 0, the frame counter value is incremented by 1 and the value has become 5. It is determined whether or not (step S11). When it becomes 5, the frame counter value is cleared to 0 (step S12), it is determined that there is no moving object, and 1 is written in the pixel selection bit register unit 29. Thereby, a frame image is selected as a JPEG compressed image (step S13). If the field counter value is not 0 as a result of the determination in step S10, the field counter value is set to 0 and the frame counter value is set to 1 (step S14). That is, when the determination that the subject is a moving object does not continue five times or more, the frequency of image switching is reduced by clearing the number of times of moving object determination to zero.
[0057]
Also, if the result of determination in step S9 is that the count value is greater than or equal to the set moving object detection sensitivity value, it is determined whether or not the frame counter value is 0 (step S15). The field counter value is incremented by 1, and it is determined whether or not the value has reached 5 (step S16). When it becomes 5, the field counter value is cleared to 0 (step S17), it is determined that there is a moving object, and 0 is written to the pixel selection bit register 29. As a result, a field image is selected as a JPEG compressed image (step S18).
[0058]
If the result of determination in step S15 is that the frame counter value is not 0, the frame counter value is set to 0 and the field counter value is set to 1 (step S19). That is, when the determination that the subject is a stationary object does not continue five or more times, the frequency of switching images is reduced by clearing the stationary object determination count to zero.
[0059]
After each process is completed, communication is performed between the system control CPU 41 and the camera control CPU 15 and the above process is repeated. By this process, the moving body detector 33 that dynamically corresponds to the gain information of the camera block 10 can dynamically switch the field image when it is determined as a moving body, and the frame image when it is determined as a stationary object. .
[0060]
Next, the overtaking scanning noise removal will be described with reference to FIG. 5 showing an example of frame memory writing state transition in the network camera 1 and FIG. 6 showing an example of frame memory reading state transition.
[0061]
In this network camera 1, the frame memory 34 has an area (area 0, area 1) for two frame images. In addition, the state of the writing of the image signal from the camera block 10 to the frame memory 34 changes in synchronization with the frame synchronization signal XDV.
[0062]
After the power is turned on, writing to area 0 is started in synchronization with the frame synchronization signal XVD. If writing of one frame is completed and area 1 is being read at the timing of the next frame synchronization signal XVD, writing to area 0 is performed. If the area 1 is not being read, the area 1 is written. That is, an area that is not read is selected as the area where the write start is performed.
[0063]
When writing to area 0, the area 0 read permission bit is set to 0 at the start of writing, and the area 0 read permission bit is set to 1 at the end of writing of one field image. That is, the read prohibition section of area 0 is from the start of writing to the end of writing of one field image.
[0064]
If the area 0 read permission state (the permission bit is 1) and JPEG compression processing is possible and the area 0 read has not ended, the area 0 read is performed. When reading of area 0 is completed or reading of area 0 is not completed and JPEG compression processing cannot be performed, an idle state is entered.
[0065]
In addition, when the reading of the 1 frame image in area 0 is completed, the area 0 read permission bit is cleared to 0. That is, the next area 0 can be read at the end of writing one field image in area 0.
[0066]
Similar processing is performed for area 1 write and read.
[0067]
If writing of one frame is completed and area 0 is being read at the timing of the next frame synchronization signal XVD, writing to area 1 is performed. If area 0 is not being read, data is written to area 0. That is, an area that is not read is selected as the area where the write start is performed.
[0068]
At the time of writing area 1, the area 1 read permission bit is set to 0 at the start of writing, and the area 1 read permission bit is set to 1 at the end of writing of one field image. That is, the read prohibition section in area 1 is from the start of writing to the end of writing of one field image.
[0069]
If the area 1 read permission state (permission bit is 1) and JPEG compression processing is possible and the area 1 read has not ended, the area 1 read is performed. When the reading of area 1 is completed or the reading of area 1 is not completed and JPEG compression processing cannot be performed, the idle state is entered.
[0070]
In addition, when the reading of one frame image in area 1 is completed, the area 1 read permission bit is cleared to 0. That is, the next area 1 can be read at the time when the writing of one field image in the area 1 is completed.
[0071]
Thus, by realizing the state transition diagrams shown in FIGS. 5 and 6, the write process does not overtake the read process in one frame image, and the overtaking scanning noise can be deleted.
[0072]
For example, as shown in FIG. 7, the network camera 1 having such a configuration is connected to a network 2 such as the Internet or a local network, and outputs an image to a plurality of PC clients 3, 4, 5 connected to the network. Is used as a camera server for attraction use and a surveillance system that distributes images in real time.
[0073]
In the embodiment of the present invention, the integrated network camera 1 in which the imaging unit 10, the signal processing unit 20, the communication interface unit 30, and the system control unit 40 are housed in one housing has been described. The network camera 1 may be constructed as a so-called head-separated camera system separately from the signal processing unit 20 or the like.
[0074]
Furthermore, an image distribution processing device that distributes image data supplied from an image data supply device such as an imaging device to a client via a network is configured by the signal processing unit 20, the communication interface unit 30, and the system control unit 40. You can also.
[0075]
【The invention's effect】
As described above, according to the present invention, when a non-interlaced frame image is generated from an interlaced image obtained by an interlaced output imaging unit, a moving body of a subject is detected by a moving body detection unit based on image data from the imaging unit. Performing detection, and determining whether the image data supplied from the imaging unit is still image data or moving image data based on the detection detection output by the moving object detection unit by the image data selection unit, Since the frame image and the field image are automatically switched depending on whether the subject is a stationary object or a moving object, a high-quality image output can be obtained.
[0076]
Further, according to the present invention, a frame memory having a storage area for a plurality of frames is used, and when writing image data to the frame memory is started, an area from which image data is not read is selected, and the frame memory is read from the frame memory. At the start of image data reading, an area where one or more fields of image data have been written is selected, and image data is written to / read from the frame memory, thereby eliminating overtaking scanning noise. Image output can be obtained.
[0077]
Further, according to the present invention, the recognition threshold of the moving object detection unit is improved by updating the difference threshold used for moving object detection in the moving object detection unit according to the setting information of the imaging unit, and the field image and the frame The image can be switched adaptively by following the setting information such as the gain value of the imaging means.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of a network camera according to the present invention.
FIG. 2 is a timing chart showing an example of memory control timing by a frame memory interface unit in the network camera.
FIG. 3 is a block diagram illustrating a configuration of a moving object detector provided in a signal processing unit in the network camera.
FIG. 4 is a flowchart showing a procedure of a frame image / field image automatic switching operation executed in the network camera.
FIG. 5 is a state transition diagram showing an example of a write state transition of a frame memory in the network camera.
FIG. 6 is a state transition diagram showing an example of a state reading of frame memory in the network camera.
FIG. 7 is a block diagram illustrating a configuration example of a network to which the network camera is connected.
FIG. 8 is a block diagram showing a configuration of a network camera used as a conventional camera server.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Network camera, 10 Camera block, 11 Interline type CCD image sensor, 12 Timing generator, 13 Front end part, 14 Digital signal processing part, 15 Camera control CPU, 16 Frame synchronous signal generation part, 17 Format conversion part, 18 Write FIFO memory, 19 frame memory write controller, 20 signal processing unit, 21 frame memory interface unit, 22 frame image frame memory read controller, 23 field image frame memory read controller, 24 image selector, 25 read FIFO memory, 26 JPEG / I / F section, 27 JPEG compression processing section, 28 CPU / I / F section, 29 image selection bit register, 30 communication interface section, 31 frame detector for moving object detection Li read controller, 32 motion detection FIFO memory, 33 motion detectors, 34 frame memory, 35 a line buffer memory, 40 system control unit

Claims (4)

Imaging means for interlaced output;
Frame image creation means and field image creation means for generating non-interlaced frame image data and field image data from the image data supplied from the imaging means;
Moving object detection means for detecting a moving object of a subject based on image data supplied from the imaging means;
Based on the detection output by the moving body detection means, it is determined whether the image data supplied from the imaging means is still image data or moving image data, and the image data supplied from the imaging means When the image data is a still image, the frame image data generated by the frame image generation unit is selected. When the image data supplied from the imaging unit is a moving image data, the frame image data is generated by the field image generation unit. Image data selecting means for selecting selected field image data;
Compression processing means for converting the image data selected by the image data selection means into compressed image data;
Network interface means for sending compressed image data obtained by the compression processing means to a network;
A frame memory having a storage area for a plurality of frames;
When starting to write image data to the frame memory, select an area where image data has not been read. When starting to read image data from the frame memory, select an area where image data has been written for one or more fields. And memory control means for writing / reading image data to / from the frame memory,
The frame image creation means generates frame image data by reading out the image data supplied from the imaging means as non-interlaced frame image data via the frame memory,
The network interface means is connected to the Internet and delivers the compressed image data in real time based on the selected image size to a plurality of clients connected to the Internet . The moving object detection unit compares the difference absolute value between the image data supplied from the imaging unit and the image data of the previous frame read out from the frame memory with a difference threshold, thereby obtaining a difference absolute value larger than the difference threshold. Comparing means for detecting a pixel that gives an image as a pixel having an image motion, counting means for counting the number of pixels detected by the comparing means, and resetting in one frame cycle, and a difference threshold for setting the difference threshold A register,
2. The network camera according to claim 1, wherein the difference threshold set in the difference threshold register is updated in accordance with setting information of the imaging means . An image distribution processing apparatus that distributes image data supplied from an interlaced output image data supply apparatus to a client via a network,
  Frame image creation means and field image creation means for generating non-interlaced frame image data and field image data from the image data supplied from the image data supply device;
  Moving object detection means for detecting a moving object of a subject based on image data supplied from the imaging means;
  Based on the detection output by the moving body detection means, it is determined whether the image data supplied from the imaging means is still image data or moving image data, and the image data supplied from the imaging means When the image data is a still image, the frame image data generated by the frame image generating means is selected, and the image data supplied from the imaging means is selected. Image data selection means for selecting the field image data generated by the field image creation means when is image data of a moving image;
  Compression processing means for converting the image data selected by the image data selection means into compressed image data;
  Network interface means for sending compressed image data obtained by the compression processing means to a network;
  A frame memory having a storage area for a plurality of frames;
  When starting to write image data to the frame memory, select an area from which image data has not been read. At the start of reading image data from the frame memory, select an area where image data has been written for one or more fields. And memory control means for writing / reading image data to / from the frame memory,
  The frame image creation means generates frame image data by reading out the image data supplied from the imaging means as non-interlaced frame image data via the frame memory,
  The network interface means is connected to the Internet and distributes the compressed image data in real time based on the selected image size to a plurality of clients connected to the Internet. The moving object detection unit compares the difference absolute value between the image data supplied from the imaging unit and the image data of the previous frame read out from the frame memory with a difference threshold, thereby obtaining a difference absolute value larger than the difference threshold. Comparing means for detecting a pixel that gives an image as a pixel having an image motion, counting means for counting the number of pixels detected by the comparing means, and resetting in one frame cycle, and a difference threshold for setting the difference threshold A register,
4. The image distribution processing apparatus according to claim 3, wherein the difference threshold set in the difference threshold register is updated in accordance with setting information of the imaging means .

Images