JP2006093784A - Image processor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processor in which a processing burden on a JPEG2000 codec can be lessened. <P>SOLUTION: When a moving body region and a fixed region are detected on a screen, a decision is made whether the moving body region and the fixed region thus detected are overlapped or not. If they are overlapped, a synthetic region is set to include a remarked region causing overlap. Subsequently, image compression is carried out by imparting an ROI effect to the set synthetic region while sustaining the compression target size of the entire image. Since it is simply required to compress the entire image by imparting the ROI effect only to the synthetic region, a processing burden on a JPEG2000 codec can be lessened. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an image processing apparatus, and more particularly to an image processing apparatus that is applied to, for example, a surveillance camera system and compresses an image signal of a screen including a region of interest.

An example of a conventional image processing apparatus of this type is disclosed in Patent Document 1. In this prior art, an image of a region of interest set in the screen is made to be a higher quality image by lowering the compression rate than images of other regions using the JPEG2000 ROI (Region Of Interest) function. Compress as follows. As a result, the data amount of the entire image can be greatly reduced while maintaining the image signal of the attention area with high image quality.
JP 2004-72655 A [H04N 1/41; G06T 1/00; G06T 7/00; H04N 1/393; H04N 5/232]

  However, in the conventional technology, when image signals of different types of attention areas are compressed with different image quality, the JPEG2000 codec must compress the images at different compression rates for each area. For this reason, there is a problem that the processing load of the JPEG2000 codec increases as the types of areas included in the screen increase.

  Therefore, a main object of the present invention is to provide an image processing apparatus capable of reducing the processing load of the JPEG2000 codec.

The invention of claim 1 is an image processing apparatus for compressing an image signal of a screen including a region of interest.
First attention area detecting means for detecting an attention area on the screen; first determination means for determining whether or not there is an overlap between the attention areas when there are a plurality of attention areas detected by the first attention area detection means; An image processing apparatus comprising: a first synthesis region setting unit that sets a synthesis region so as to include overlapping regions of interest when the determination result by the first determination unit is affirmative.

  In the first aspect of the present invention, when a plurality of attention areas are detected by the first attention area detection means on the screen including the attention area, the first determination means determines whether or not there is an overlap between the attention areas. To do. As a result, when it is determined that there is an overlap, the composite region is set by the first composite region setting means so as to include the overlapped attention region. In this case, since the synthesis area is set so as to include the attention area determined to be overlapped, the attention area can be reduced. Therefore, it is possible to reduce the processing load of the image compression means for compressing the entire screen by giving the ROI effect to the partial image in the composite area.

  The invention of claim 2 is dependent on claim 1, and after setting the synthesis area by the first synthesis area setting means, the attention area is detected by the second attention area detection means and the second attention area detection means for detecting the attention area of the screen. First peripheral area specifying means for specifying an area for which no composite area has been set as a peripheral area when the area is not detected, importance determined for each of the composite area and the peripheral area so that the compression target size of the screen is constant First calculation means for calculating compression ratios of the synthesis area and the peripheral area based on the coefficients, respectively, and compressing the partial image signals of the synthesis area and the peripheral area based on the compression ratio obtained by the first calculation means, respectively. An image processing apparatus further comprising first image compression means.

  In the invention of claim 2, when the attention area is not detected after the synthesis area is set, the area of the screen in which the synthesis area is not set is specified as the peripheral area. Next, the compression ratios of the composite area and the peripheral area are respectively calculated and calculated based on the importance coefficients determined for the composite area and the peripheral area so that the compression target size of the screen becomes constant for each screen. Then, the partial image signals in the composite area and the peripheral area are compressed based on the obtained compression rate. In this case, the image compression can be performed with the ROI effect applied to the partial image signal in the composite area while maintaining the compression target size of the screen constant.

  The invention of claim 3 is dependent on claim 2, and when the attention area is detected by the second attention area detection means, the first peripheral area specifying means surrounds the area where the attention area and the synthesis area are not detected. A second surrounding area specifying unit that specifies the area, and the first calculating unit combines based on the importance coefficient determined for each of the combining area, the attention area, and the surrounding area so that the compression target size of the screen is constant. Second calculation means for calculating the compression ratio of the region, the attention area and the peripheral area by calculation, respectively, and the first image compression means is based on the compression ratio determined by the second calculation means, An image processing apparatus includes second image compression means for compressing each partial image signal of a region.

  According to the third aspect of the present invention, when the attention area is detected after setting the synthesis area, the area of the screen in which the attention area and the synthesis area are not set is specified as the peripheral area. Next, calculate the compression ratio of the attention area, composition area, and surrounding area based on the importance factor defined for each attention area, composition area, and surrounding area so that the compression target size of the screen is constant for each screen. And ask. Then, the partial image signals of the attention area, the synthesis area, and the peripheral area are respectively compressed based on the obtained compression rate. In this case, image compression can be performed with the ROI effect applied to the partial image signals in the attention area and the synthesis area while maintaining the compression target size of the screen constant.

The invention of claim 4 is dependent on claim 1 and, when the determination result by the first determination means is negative, third peripheral area specifying means for specifying an area in which the attention area is not detected as a peripheral area,
Third calculation means for calculating the respective compression rates of the attention area and the peripheral area based on the importance coefficient determined for each of the attention area and the peripheral area, and attention based on the compression ratio obtained by the third calculation means The image processing apparatus further includes third image compression means for compressing each partial image signal in the region and the peripheral region.

  In the invention of claim 4, when the attention area determined to have no overlap is included on the screen, the compression of each of the attention area and the surrounding area is performed based on the importance coefficient determined for each of the attention area and the surrounding area. The rate is calculated. Then, each partial image signal in the attention area and the peripheral area is compressed based on the obtained compression rate. In this case, even when a region of interest that is not set as a synthesis region because there is no overlap is included on the screen, image compression is performed with a ROI effect on the partial image signal of the region of interest while maintaining the compression target size of the screen constant. can do.

  The invention according to claim 5 is dependent on any one of claims 1 to 4, wherein the attention area and the synthesis area are both rectangular, and the X coordinate and Y coordinates of the two vertices in each diagonal direction of the overlapping attention area Coordinate detecting means for detecting coordinates, and coordinate specifying means for specifying the maximum and minimum values of the X coordinates and the maximum and minimum values of the Y coordinates detected by the coordinate detecting means, further comprising: Second composite region setting for setting a composite region having a vertex specified by the minimum value of the X coordinate and the Y coordinate specified by the coordinate specifying means and a vertex specified by the maximum value of the X coordinate and the Y coordinate in the diagonal direction An image processing apparatus including means.

  In the invention of claim 5, since the shapes of the attention area and the synthesis area are both rectangular, if the coordinates of the two vertices in the respective diagonal directions can be identified, the shapes of the attention area and the synthesis area can be identified. For this reason, the X coordinate and the Y coordinate of the two vertices in the diagonal direction of the overlapping attention area are detected. Next, the maximum value and the minimum value are specified from the detected X coordinates, and the maximum value and the minimum value are specified from the detected Y coordinates. And the synthetic | combination area | region which has a vertex specified by the minimum value of the specified X coordinate and Y coordinate and a vertex specified by the maximum value of X coordinate and Y coordinate in a diagonal direction is set. In this case, the synthesis area can be easily set from the attention area where the overlap is detected.

  The invention of claim 6 is dependent on any one of claims 1 to 5, wherein the first attention area detecting means is a fixed area detecting means for detecting a fixed area whose position is fixed, and the movement of the subject. A moving body region detecting unit for detecting a moving body region whose position moves; a first determining unit including a second determining unit for determining an overlap between the fixed region and the moving body region; and a first synthesis region setting unit including: 2 is an image processing apparatus including a third synthesis area setting unit that sets a synthesis area so as to include the fixed area and the moving object area when it is determined by the determination unit that the fixed area and the moving object area overlap.

  In the invention of claim 6, first, a fixed area whose position is fixed on the screen and a moving body area whose position moves in accordance with the movement of the subject are detected. Next, it is determined whether or not there is an overlap between the detected fixed area and the moving object area. As a result, when it is determined that there is an overlap, the synthesis area is set so as to include the fixed area and the moving object area. In this case, when the fixed area and the moving object area overlap, an area including the overlapping fixed area and moving object area can be set as a composite area.

  A seventh aspect of the invention is an image processing apparatus according to any one of the first to sixth aspects, wherein the image signal on the screen is an image signal output from a surveillance camera. In this case, when the image signal output from the surveillance camera is also compressed with the ROI effect, the processing load of the image compression means can be reduced by setting the synthesis area.

  According to the present invention, when a plurality of attention areas are overlapped, one composite area can be set so as to include the attention areas where the overlap occurs, so that the processing burden of each image compression unit can be reduced. Can do.

  The above object, other objects, features and advantages of the present invention will become more apparent from the following detailed description of embodiments with reference to the drawings.

  Referring to FIG. 1, surveillance camera system 10 according to an embodiment of the present invention includes surveillance camera 12, hard disk recorder 14, and monitor 40. The surveillance camera system 10 can compress an image of a subject photographed by the surveillance camera 12 and record it on the hard disk recorder 14, or decompress the recorded compressed image and reproduce it on the monitor 40.

  The surveillance camera 12 provides the hard disk recorder 14 with an image of the subject as an analog signal every field period in order to monitor whether or not the subject is moving.

  In the hard disk recorder 14, a DI / F 16, a CPU 20, a JPEG2000 codec 28, a memory control circuit 24, an HDD-I / F 30, a flash memory 36, and a video output circuit 38 are connected to each other via a bus 22. The motion detection circuit 18 is connected to the DI / F 16 and the CPU 20, the HDD 32 containing the hard disk 34 is connected to the HDD-I / F 30, and the SDRAM 26 is connected to the memory control circuit 24.

  The CPU 20 instructs the DI / F 16 to take in the analog signal of the subject given at a predetermined interval from the monitoring camera 12. The D-I / F 16 first converts the captured analog signal by a video decoder (not shown) inside the D / I / F 16 into a Y signal that is a luminance signal and a U (R−Y) signal that is a color difference signal. Conversion to V (BY) signal. Next, the converted Y signal, U signal, and V signal are converted into digital image signals by an A / D conversion circuit (not shown) in the D-I / F 16 and the converted image signals are converted. Is supplied to the motion detection circuit 18 and the memory control circuit 24.

  The motion detection circuit 18 extracts a Y signal, which is a luminance signal, from the image signal given from the DI / F 16 and determines the magnitude of the Y signal of the current screen and the magnitude of the Y signal of the screen one field before. The luminance change is obtained by comparison. The luminance change of the Y signal is obtained for each block on the screen of the monitor 40 divided into a plurality of blocks. The obtained luminance change of the Y signal is given to the CPU 20.

  The CPU 20 determines for each block whether or not the luminance change of the given Y signal has exceeded a preset detection threshold. As a result, when there is a block in which a Y signal having a luminance change exceeding the detection threshold is detected, it is determined that the subject has moved in that block. As described above, the internal alarm is detected when the CPU 20 determines that the luminance change of the Y signal obtained by the motion detection circuit 18 has exceeded the detection threshold. When an internal alarm is detected in a certain block, the CPU 20 identifies the block, and sets a moving object region, which is one of the regions of interest, using the ROI function of JPEG2000 for the identified block. The moving body region set in this way moves on the screen in accordance with the movement of the subject. The image of the subject in the moving object region is compressed so as to be a higher quality image than the image in the peripheral region that is the region excluding the attention region on the screen.

  In addition, a fixed area which is another attention area using the ROI function is set on the screen of the monitor 40. Since the position of the fixed area is fixed in the screen, no internal alarm is installed in the blocks constituting the fixed area. For this reason, the image in the fixed area is compressed so as to be a higher quality image than the image in the peripheral area regardless of the movement of the subject. On the other hand, the image signal given from the D-I / F 16 to the memory control circuit 24 is written in the SDRAM 26.

  The importance coefficient and the image compression calculation formula determined for each area type are stored in the flash memory 36. Here, the importance coefficient is a constant related to the image quality of the image after compression, and the higher the value, the higher the compression rate of the area, so the image quality of the image after compression becomes worse. In addition, the compression ratio of each region can be obtained based on the importance coefficient so that the compression target size of the entire screen is constant by the image compression calculation formula. As will be described later, there are a plurality of types of calculation formulas for image compression in accordance with the types of areas included in one screen.

  The CPU 20 gives an image signal compression command to the JPEG2000 codec 28. When the JPEG 2000 codec 28 receives the compression command for the image signal, it requests the memory control circuit 24 to read the image signal. Next, the JPEG 2000 codec 28 takes in the image signal read from the SDRAM 26 by the memory control circuit 24. Then, the CPU 20 detects the type of area and the image size of each area included in the pre-compression screen, and based on the detected type of area and the image size of each area, the importance coefficient and The image compression calculation formula is read out, and the read importance coefficient and the image compression calculation formula are given to the JPEG2000 codec 28.

  The JPEG 2000 codec 28 compresses the image signal for each region based on the importance coefficient given from the CPU 20 and the image compression calculation formula. That is, the JPEG 2000 codec 28 calculates the compression rate of each area so that the compression target size of the entire screen is constant based on the importance coefficient given by the CPU 20 and the image compression calculation formula. Next, the JPEG 2000 codec 28 compresses the image signal of each area at the obtained compression rate. As a result, the image quality of the compressed image varies depending on the type of region, and the compression target size of the entire screen is constant. The JPEG 2000 codec 28 generates a compressed image signal by compressing the image signal, and then requests the memory control circuit 24 to write the generated compressed image signal. The memory control circuit 24 writes the compressed image signal into the SDRAM 26 in response to a request from the JPEG2000 codec 28.

  Next, the CPU 20 gives a recording command of the compressed image signal to the HDD-I / F 30. The HDD-I / F 30 requests the memory control circuit 24 to extract a compressed image signal in accordance with a recording command, and provides the compressed image signal extracted from the SDRAM 26 by the memory control circuit 24 to the HDD 32. The HDD 32 records the given compressed image signal on the hard disk 34 in a file format. In the hard disk 34, the recorded compressed image signal files are managed in the order of shooting.

  Next, a case where the compressed image signal recorded on the hard disk 34 is reproduced will be described. First, the CPU 20 instructs the HDD-I / F 30 to read out the compressed image signal. Receiving the read command, the HDD-I / F 30 controls the HDD 32 to sequentially read out the compressed image signals corresponding to the subject photographed by the monitoring camera 12 from the hard disk 34 in the order of photographing. Then, the CPU 20 instructs the memory control circuit 24 to write the read compressed image signal to the SDRAM 26. Receiving the write command, the memory control circuit 24 writes the compressed image signal into the SDRAM 26.

  Next, the CPU 20 gives a decompression command for the compressed image signal to the JPEG2000 codec 28. Upon receiving the compressed image signal decompression command, the JPEG 2000 codec 28 requests the memory control circuit 24 to read the compressed image signal, and takes in the compressed image signal written in the SDRAM 26. The captured compressed image signal is expanded according to a method defined in JPEG2000. When a region of interest is set in the compressed image signal to be decompressed, the image signal of the region of interest and the image signal of the peripheral region are respectively read from the flash memory 36 with importance setting constants and image compression values for each region. Expand based on the formula. The decompressed image signal decompressed by the JPEG2000 codec 28 is given to the memory control circuit 24 and written into the SDRAM 26 by the memory control circuit 24.

  Further, the CPU 20 gives a processing instruction for the decompressed image signal to the video output circuit 38. The video output circuit 38 that has received the processing instruction for the decompressed image signal requests the memory control circuit 24 to extract the decompressed image signal for each field period, and the decompressed image signal extracted from the SDRAM 26 by the memory control circuit 24. receive.

  The video output circuit 38 encodes the received decompressed image signal into a composite image signal, and displays the encoded composite image signal on the screen of the monitor 40. At this time, the image of the block in the region of interest has a higher image quality than the image of the peripheral region due to the ROI effect.

  Next, with reference to FIGS. 2A to 3B, a case where a plurality of attention areas exist on the screen of the monitor 40 will be described. First, referring to FIG. 2A, the screen of the monitor 40 is divided into 64 blocks (8 × 8 blocks). On this screen, a rectangular fixed area composed of 3 × 3 blocks whose positions are fixed is provided. Internal alarms are installed in blocks other than the blocks constituting the fixed area.

  With reference to FIG. 2 (B), the case where a moving body area | region appears on a screen is demonstrated. The moving object area is a kind of attention area and is composed of blocks in which an internal alarm is detected. The moving object area has a rectangular shape composed of 2 × 3 blocks, and moves on the screen as the subject moves.

  Next, with reference to FIG. 3 (A), a case where the moving object region has moved and the moving object region and the fixed region partially overlap will be described. Since the importance factor of the moving object region is set to a value different from the importance factor of the fixed region, the JPEG 2000 codec 28 compresses the image signal of the moving object region at a compression rate different from that of the image signal of the fixed region. As described above, when the types of attention areas increase, the JPEG2000 codec 28 has to compress the image signals of the respective areas at different compression rates, and thus the processing load increases. Therefore, as shown in FIG. 3B, a rectangular composite area is set as a new area including the moving object area and the fixed area. As a result, rather than reading out the importance factor of the moving object region and the importance factor of the fixed region, and compressing the moving object region and the fixed region at different compression rates, the importance factor of the synthesis region is set in advance. The processing load on the JPEG 2000 codec 28 can be reduced by compressing the image in the synthesis area at the compression rate obtained based on the set importance coefficient.

  Next, what has been qualitatively described with reference to FIGS. 2A to 3B will be quantitatively described using an image compression calculation formula. First, the criteria for determining the positional relationship between the fixed area and the moving object area will be described. Here, it is assumed that both the fixed area and the moving object area are rectangular. Therefore, the positions of the fixed area and the moving object area on the screen can be represented by the coordinates of the two vertices in the respective diagonal directions. Accordingly, the coordinates of the vertices A1 and A2 of the fixed area are (Xa1, Ya1) and (Xa2, Ya2), respectively. At this time, it is assumed that the coordinates of the vertices A1 and A2 have a relationship of Xa1 <Xa2 and Ya1 <Ya2. In addition, the coordinates of the vertexes B1 and B2 of the moving object region are (Xb1, Yb1) and (Xb2, Yb2), respectively. At this time, it is assumed that the coordinates of the vertices B1 and B2 have a relationship of Xb1 <Xb2 and Yb1 <Yb2.

  As shown in FIGS. 4A to 5B, the positional relationship between the fixed area and the moving object area is such that when the fixed area and the moving object area are separated from each other, a part of the moving object area is a part of the fixed area. In the case where the fixed area is completely included in the moving area, the moving area may be completely included in the fixed area.

  The positional relationship of the coordinates of each vertex when the fixed area and the moving object area are separated from each other will be described. In this case, as shown in FIG. 4A, the positional relationship between the fixed area and the moving object area is such that the moving object area is located away from the upper, lower, left, and right positions of the fixed area. Either.

  First, in order for the moving object region to be separated upward in the fixed region, the condition shown in Equation 1 must be satisfied.

[Equation 1]
Ya2 <Yb1
In order for the moving body region to be separated downward in the fixed region, the condition shown in Equation 2 must be satisfied.

[Equation 2]
Yb2 <Ya1
In order for the moving object region to be separated in the left direction of the fixed region, the condition shown in Equation 3 must be satisfied.

[Equation 3]
Xb2 <Xa1
In order for the moving object region to be separated in the right direction of the fixed region, the condition shown in Equation 4 must be satisfied.

[Equation 4]
Xa2 <Xb1
Next, referring to FIG. 4B, the position of the coordinates of each vertex when a part of the fixed area overlaps with a part of the moving object area, that is, when a part of the moving object area is included in the fixed area The relationship will be described.

  First, in order for the lower right part of the moving object region to be included in the fixed region, the condition shown in Equation 5 must be satisfied.

[Equation 5]
Xa1 ≦ Xb2 ≦ Xa2 and Ya1 ≦ Yb1 ≦ Ya2
In order for the upper right part of the moving object region to be included in the fixed region, the condition shown in Equation 6 must be satisfied.

[Equation 6]
Xa1 ≦ Xb2 ≦ Xa2 and Ya1 ≦ Yb2 ≦ Ya2
In order for the lower left part of the moving object region to be included in the fixed region, the condition shown in Equation 7 must be satisfied.

[Equation 7]
Xa1 ≦ Xb1 ≦ Xa2 and Ya1 ≦ Yb1 ≦ Ya2
In order for the upper left part of the moving object region to be included in the fixed region, the condition shown in Equation 8 must be satisfied.

[Equation 8]
Xa1 ≦ Xb1 ≦ Xa2 and Ya1 ≦ Yb2 ≦ Ya2
In addition, the positional relationship between the coordinates of each vertex when the fixed area is completely included in the moving object area will be described. In this case, the positional relationship between the fixed region and the moving object region must satisfy the condition shown in Equation 9 as can be seen from FIG.

[Equation 9]
Xb1 ≦ Xa1 ≦ Xb2 and Xb1 ≦ Xa2 ≦ Xb2 and Yb1 ≦ Ya1 ≦ Yb2 and Yb1 ≦ Ya2 ≦ Yb2
Furthermore, the positional relationship between the coordinates of each vertex when the moving object region is completely included in the fixed region will be described. In this case, the positional relationship between the fixed area and the moving object area must satisfy the condition shown in Equation 10 as can be seen from FIG.

[Equation 10]
Xa1 ≦ Xb1 ≦ Xa2 and Xa1 ≦ Xb2 ≦ Xa2 and Ya1 ≦ Yb1 ≦ Ya2 and Ya1 ≦ Yb2 ≦ Ya2
Next, as shown in FIG. 6A, a case will be described in which when a part of the moving object region is included in the fixed region, a composite region including the fixed region and the moving object region is newly set. In this case, as shown in FIG. 6B, the minimum value and the maximum value among the four X coordinates Xa1, Xa2, Xb1, and Xb2 of the fixed region and the moving object region are respectively Xmin and Xmax, and similarly, the Y coordinate Ya1 , Ya2, Yb1, and Yb2 are Ymin and Ymax, respectively, and the coordinates of the two vertices C1 and C2 in the diagonal direction of the composite region are (Xmin, Ymin) and (Xmax, Ymax), respectively. ) As a result, the fixed area and the moving object area are completely included in the newly set synthesis area.

  Similarly, a composite area is newly set when the fixed area is completely included in the moving object area and when the moving object area is completely included in the fixed area. In this case, the newly set synthesis area matches the moving object area and the fixed area, respectively.

  Next, a method for compressing an image with the ROI effect without changing the compression target size in each state shown in FIGS. 2A to 3B will be described. The image size of the entire screen is W, and the compression target size is Q.

First, a case where the entire screen is a peripheral area, that is, a case where no attention area is set will be described. At this time, the compression target size Q is
[Equation 11]
Q = Lz * k1 * Wz
Where Wz: image size of the peripheral area
Lz: Importance coefficient of surrounding area
k1: First proportional constant Next, as shown in FIG. 2A, the compression target size Q when only a fixed area exists in the screen is
[Equation 12]
Q = (La * Wa + Lz * Wz) * k2
Where Wa: image size of the fixed area
La: Importance coefficient of fixed area
k2: Second proportional constant Using the fact that the compression target sizes Q of Equations 11 and 12 are equal, Equation 12 is as follows.

[Equation 13]
Q = (La * Wa + Lz * Wz) * k2
= (La * Wa + Lz * Wz) * (Wz * Lz / (Wa * La + Wz * Lz)) * k1
= ((La * Wa) * (Wz * Lz / (Wa * La + Wz * Lz)) * k1 + ((Lz *
Wz) * (Wz * Lz / (Wa * La + Wz * Lz)) * k1
It becomes. As can be seen from Equation 13, by replacing the second proportional constant k2 with (Wz * Lz / (Wa * La + Wz * Lz)) * k1, the compression target size Q of the entire screen is not changed, and the fixed region and The image signal of the entire screen including the peripheral area can be compressed.

Here, the case where only the fixed area exists in the screen has been described, but it can also be obtained in the same manner when only the moving object area exists in the screen. That is,
[Formula 14]
Q = (Lb * Wb + Lz * Wz) * k3
= ((Lb * Wb) * (Wz * Lz / (Wb * Lb + Wz * Lz)) * k1 + ((Lz *
Wz) * (Wz * Lz / (Wb * Lb + Wz * Lz)) * k1
Where Wb: image size of moving object area
Lb: Importance coefficient of moving object region
k3: a third proportionality constant. As can be seen from Equation 14, the third proportionality constant k3 is replaced with (Wz * Lz / (Wb * Lb + Wz * Lz)) * k1, so that the moving object region and the moving object region are changed without changing the compression target size Q of the entire screen. The image signal of the entire screen including the peripheral area can be compressed.

Next, as shown in FIG. 2 (B), there are a fixed area and a moving object area in the screen.
[Equation 15]
Q = (La * Wa + Lb * Wb + Lz * Wz) * k4
= ((La * Wa) * (Wz * Lz / (Wa * La + Wb * Lb + Wz * Lz)) *
k1 + ((Lb * Wb) * (Wz * Lz / (Wa * La + Wb * Lb + Wz * Lz)))
* K1 + ((Lz * Wz) * (Wz * Lz / (Wa * La + Wb * Lb + Wz * Lz))
* K1
Here, k4: Fourth proportionality constant As can be seen from Equation 15, the compression target size Q of the entire screen is replaced by replacing the fourth proportionality constant k4 with (Wz * Lz / (Wa * La + Wb * Lb + Wz * Lz)). The image signal of the entire screen including the fixed area, the moving object area, and the peripheral area can be compressed without changing.

As shown in FIG. 3B, when at least a part of the moving object region is included in the fixed region, the composite region is formed so as to completely include the fixed region and the moving object region. At this time, if the compression target size is obtained in the same manner,
[Equation 16]
Q = (Lc * Wc + Lz * Wz) * k5
= ((Lc * Wc) * (Wz * Lz / (Wc * Lc + Wz * Lz)) * k1 + ((Lz *
Wz) * (Wz * Lz / (Wc * Lc + Wz * Lz)) * k1
Where Wc: image size of the composite area
Lc: Importance coefficient of the synthesis area
k5: The fifth proportionality constant. As can be seen from Equation 16, when the composite area is set from the fixed area and the moving object area, the fifth proportional constant k5 is replaced with (Wz * Lz / (Wc * Lc + Wz * Lz)) * k1 to obtain the entire screen. Without changing the compression target size Q, the image signal of the entire screen including the composite area and the peripheral area can be compressed. By newly setting the synthesis area in this way, it is not necessary to compress the fixed area and the moving object area at different compression rates, so that the processing load of the JPEG2000 codec 28 can be reduced.

  Next, referring to FIG. 7 and FIG. 8, in this surveillance camera system 10, when the attention area is set in a part of the screen, the image signal of each area is obtained without changing the compression target size of the entire screen. A processing flow for compression will be described.

  First, in step S1, the entire screen of the monitor 40 is divided into 8 × 8 blocks. In step S3, the arrangement of the block for setting the internal alarm on the divided screen, that is, the block for obtaining the luminance change of the Y signal by the motion detection circuit 18 is determined. In step S5, a detection threshold value for the luminance change of the Y signal detected by the motion detection circuit 18 is set and stored in the flash memory 36. In step S <b> 7, importance coefficients are set according to the peripheral area, fixed area, moving object area, and composite area, and stored in the flash memory 36. In step S9, formulas 11, 13, 13, 15, and 16 for compression without changing the compression target size of the image signal of the entire screen are stored in the flash memory 36.

  In step S11, the process waits until a vertical synchronization signal is generated. When the vertical synchronization signal is generated, in step S13, the D / I 16 is instructed to take in an analog signal from the monitoring camera 12. The D-I / F 16 generates a Y signal, a U signal, and a V signal based on the acquired analog signal, and converts the generated image signal into a digital signal.

  In step S15, it is determined whether there is a block in which no internal alarm is installed. As a result, if NO is determined, the process proceeds to step S17, and it is determined whether there is a block whose movement is detected by the internal alarm. As a result, if NO is determined, the process further proceeds to step S19 to determine whether or not the screen is composed of only the peripheral area.

  In step S19, if NO is determined, the process returns to step S15, and if YES is determined, the process proceeds to step S21. In step S21, the importance coefficient Lz and Equation 11 of the peripheral area are read from the flash memory 36, and the process proceeds to step S45 described later.

  If YES is determined in step S17, in step S23, a moving object region including a block in which an internal alarm is detected is specified. In step S25, the importance factors Lb and Lz and Equation 14 of the moving object region and the peripheral region are read from the flash memory 36, and the process proceeds to step S45 described later.

  If YES is determined in step S15, in step S27, a fixed area composed of blocks in which no internal alarm is installed is specified. In step S29, it is determined whether there is a block whose movement is detected by an internal alarm. As a result, when it is determined YES, the process proceeds to step S31, and a moving object region constituted by blocks in which an internal alarm is detected is specified.

  In step S33, it is determined whether or not there is an overlap between the fixed area and the moving object area. Here, whether or not there is an overlap is determined by whether or not the coordinates of the vertices of the fixed area and the moving object area satisfy the conditions shown in any one of Formulas 5 to 8. In step S35, the importance coefficients La, Lb, Lz and Equation 15 of the fixed area, moving object area, and peripheral area are read from the flash memory 36, and the process proceeds to step S45 described later.

  If YES is determined in step S33, in step S37, the coordinates of the vertices in the diagonal direction of the moving object region and the fixed region are detected. In step S39, a composite area including the moving object area and the fixed area is set. That is, of the detected coordinates of each vertex, a point determined by the maximum value in the X direction and the maximum value in the Y direction, and a point determined by the minimum value in the X direction and the minimum value in the Y direction are the vertices in the diagonal direction. A rectangular composite area is set. In step S41, the importance coefficients Lc and Lz and Equation 16 of the composite area and the peripheral area are read from the flash memory 36, and the process proceeds to step S45 described later.

  If NO is determined in step S29, in step S43, the importance coefficients La and Lz and the number 13 of the fixed area and the peripheral area are read from the flash memory 36, and the process proceeds to step S45 described later.

  In step S45, the JPEG2000 codec 28 obtains a compression rate for each area based on the importance coefficient for each area given from the CPU 20 and an image compression calculation formula determined according to the type of area included in the screen. The image signal of each region is compressed with the compression ratio obtained. As a result, it is possible to compress the image of the attention area or the synthesis area so that the compression target size is constant while maintaining a higher image quality than the image of the peripheral area. The compressed image signal is recorded on the hard disk 34 via the HDD 32. In step S47, it is determined whether or not recording on the hard disk 32 is to be terminated. As a result, if NO is determined, the process returns to step S11, and if YES is determined, the process is terminated.

  In the above-described embodiment, the luminance change of the Y signal is obtained for each block, and the movement of the subject is detected when the obtained luminance change is larger than the detection threshold. However, the method of detecting the motion of the subject is not limited to this, and a motion vector may be obtained for each block, and the motion of the subject may be detected based on the obtained magnitude of the motion vector.

  The composition area can be set not only when the fixed area and the moving body area overlap, but also when the fixed area and the fixed area overlap, or when the moving body area and the moving body area overlap.

  In addition, the setting of the synthesis region is not limited to the case where two attention regions overlap, and may be a case where three or more attention regions overlap.

  As can be understood from the above description, when a plurality of attention areas are detected on the screen including the attention area, it is determined whether or not there is an overlap between the attention areas. As a result, when it is determined that there is an overlap, the synthesis region is set so as to include the attention region where the overlap has occurred. In this case, since the synthesis area is set so as to include the attention area determined to be overlapped, the attention area can be reduced. Therefore, when the entire image is compressed with the ROI effect applied to the partial image in the composite area, the processing load of the JPEG2000 codec can be reduced.

It is a block diagram which shows one Example of this invention. It is an illustration figure which shows a part of operation | movement of FIG. 1 Example. It is an illustration figure which shows a part of other operation | movement of FIG. 1 Example. It is an illustration figure which shows the other part of operation | movement of FIG. 1 Example. FIG. 10 is an illustrative view showing yet another portion of the operation of the embodiment in FIG. 1; It is an illustration figure which shows a part of other operation | movement of FIG. 1 Example. It is a flowchart which shows the others of operation | movement of FIG. 1 Example. It is a flowchart which shows a part of operation | movement of FIG. 1 Example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Surveillance camera system 14 ... Hard disk recorder 18 ... Motion detection circuit 20 ... CPU
28 ... JPEG2000 codec

Claims (7)

In an image processing apparatus that compresses an image signal of a screen including a region of interest,
First attention area detection means for detecting the attention area of the screen;
First determination means for determining whether or not there is an overlap between the attention areas when there are a plurality of the attention areas detected by the first attention area detection means, and the determination result by the first determination means is positive An image processing apparatus comprising: a first synthesis area setting unit that sets a synthesis area so as to include overlapping attention areas. Second attention area detection means for detecting the attention area of the screen after the synthesis area setting by the first synthesis area setting means;
First peripheral area specifying means for specifying, as a peripheral area, an area in which the composite area is not set when the attention area is not detected by the second attention area detecting means;
First calculation means for calculating compression ratios of the synthesis area and the peripheral area based on importance factors determined for the synthesis area and the peripheral area so that the compression target size of the screen is constant; The image processing apparatus according to claim 1, further comprising a first image compression unit that compresses each partial image signal of the synthesis region and the peripheral region based on a compression rate obtained by the first calculation unit. When the attention area is detected by the second attention area detection means,
The first peripheral area specifying means includes second peripheral area specifying means for specifying, as a peripheral area, an area in which the attention area and the synthesis area are not detected,
The first calculation means is configured to determine the combination area, the attention area, and the attention area based on importance coefficients determined for the combination area, the attention area, and the peripheral area so that a compression target size of the screen is constant. Second calculation means for calculating the compression ratio of the peripheral area by calculation, respectively, and the first image compression means is configured to calculate the combination area, the attention area, and the peripheral area based on the compression ratio calculated by the second calculation means. The image processing apparatus according to claim 2, further comprising second image compression means for compressing each partial image signal of the region. When the determination result by the first determination means is negative,
Third peripheral region specifying means for specifying, as a peripheral region, a region where the attention region has not been detected;
Third calculation means for calculating the compression ratio of the attention area and the peripheral area based on the importance coefficient determined for each of the attention area and the peripheral area, and based on the compression ratio determined by the third calculation means The image processing apparatus according to claim 1, further comprising third image compression means for compressing each partial image signal of the region of interest and the peripheral region. The attention area and the synthesis area are both rectangular.
Coordinate detection means for detecting X and Y coordinates of two vertices in the respective diagonal directions for the overlapped attention area, and maximum and minimum values of the X coordinate detected by the coordinate detection means, and the Y coordinate Further comprising coordinate specifying means for specifying the maximum value and the minimum value of
The first synthesis area setting means diagonally specifies the vertex specified by the minimum value of the X coordinate and the Y coordinate specified by the coordinate specifying means and the vertex specified by the maximum value of the X coordinate and the Y coordinate. The image processing apparatus according to claim 1, further comprising: a second composite area setting unit that sets the composite area having a direction. The first region of interest detection means includes a fixed area detection means for detecting a fixed area whose position is fixed, and a moving body area detection means for detecting a moving body area whose position moves with the movement of the subject,
The first determination means includes second determination means for determining an overlap between the fixed area and the moving object area, and the first synthesis area setting means is configured to determine the fixed area and the moving object area by the second determination means. 6. The image processing apparatus according to claim 1, further comprising: a third synthesis area setting unit that sets a synthesis area so as to include the fixed area and the moving object area when it is determined that there is an overlap with the fixed area. .   The image processing apparatus according to claim 1, wherein the image signal of the screen is an image signal output from a surveillance camera.

Images