JP2008086040A - Motion detecting unit - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a motion detecting unit whose motion detection accuracy can be improved. <P>SOLUTION: When an exposure control operation functions to change shutter speed or gains of an electronic shutter, brightness value of a background image can be kept constant by normalizing the brightness value of the image with the shutter speed or gains using a wide-D range image generating part 3. Accordingly, the background image can be protected from generating frame differences, and the motion detection accuracy can be improved as compared with a method of, for example, detecting motion of an object by using an image taken by an imaging part 2 directly. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a motion detection device that captures an image with an imaging unit each time a predetermined time elapses and detects a motion of a subject by detecting different portions from the captured images.

  Conventionally, this type of technology includes, for example, a security system using a camera and a surveillance camera. In such a monitoring camera or the like, first, an image at a predetermined position is captured every time a certain time elapses, and the difference in luminance value between the previous frame and the current frame is calculated at each pixel of the captured image. To do. Then, by detecting a portion where the difference is large, the movement of a specific subject is detected in distinction from the background without movement (see, for example, Patent Document 1).

  By the way, in this type of surveillance camera or the like, exposure control is generally performed so that the brightness of an image to be captured falls within a certain range. In such exposure control, first, the brightness of the image of the previous frame is digitized to obtain a brightness evaluation value. Then, the evaluation value is compared with a reference value (exposure target value) of brightness set from the outside, and the shutter speed of the electronic shutter and the amplification factor of the output of the image sensor are controlled so that the two values substantially coincide. . Thereby, the brightness of the whole image is made to substantially coincide with the exposure target value regardless of the luminance change of the light source.

JP-A-9-32222028

However, in the above conventional technique, when the luminance change occurs in the light source, the exposure control operation works, and the brightness of the entire image slightly changes even when the shutter speed and the amplification factor are controlled. Therefore, there is a problem that the luminance value of the entire image changes, a frame difference occurs in the entire image, and the movement of the subject is erroneously detected even when the subject does not actually move.
Even when the subject moves, if the brightness evaluation value changes due to the motion of the subject, the exposure control operation works, and the shutter speed and the amplification factor change. As a result, the luminance value of the background image changes, a frame difference occurs in the background image, and it is erroneously determined that there is motion in the background portion. As a result, the accuracy of motion detection decreases. It was.

  An object of the present invention is to solve an unsolved problem of the conventional motion detection device, and to provide a motion detection device capable of improving the accuracy of motion detection.

  In order to solve the above problems, a motion detection device of the present invention is a motion detection device that detects a motion of a subject by detecting a portion where a luminance value has changed between a plurality of images taken at different times. Normalization means for normalizing the brightness value of the image at least one of a shutter speed of the electronic shutter at the time of capturing the image and a gain multiplied by the output of the image sensor that captured the image, and the normalization means And detecting means for detecting a portion where the brightness value has changed from between the images whose brightness values are normalized.

Further, the normalizing means may multiply the luminance value of the image by a conversion coefficient calculated by dividing a multiplication value of a maximum shutter speed and a maximum gain by the shutter speed and the gain. .
According to such a configuration, for example, when the exposure control operation is performed and the shutter speed of the electronic shutter or the amplification factor of the sensor output is changed, the luminance value of the image is normalized by the shutter speed or the amplification factor. The brightness value of the image can be kept constant. Therefore, it is possible to prevent a frame difference from occurring in the background image, and for example, it is possible to improve the accuracy of motion detection compared to a method of detecting the motion of the subject using the image captured by the imaging unit as it is. .

Hereinafter, an embodiment of a motion detection device according to the present invention will be described with reference to the drawings.
The motion detection device 1 detects the motion of a subject by detecting different portions from each other between images captured every time a predetermined time passes by an imaging unit capable of capturing a predetermined image. At this time, the motion detection apparatus corrects the luminance value of the image used for the detection of the different portions by the control parameter of the exposure control executed by the imaging unit. Therefore, the luminance value of the background image can be kept constant, frame differences can be prevented from occurring in the background image, and motion detection accuracy can be improved.

<Configuration of motion detection device>
FIG. 1 is a block diagram showing an internal configuration of an embodiment of the present invention. As shown in FIG. 1, the motion detection device 1 includes an imaging unit 2, a wide D range image generation unit 3, and a motion detection unit 4. As shown in FIG. 2, the imaging unit 2 includes an imaging lens 5, an analog / digital video signal processing main line 6, and an exposure control unit 7.

The imaging lens 5 is disposed above the image sensor 8 with the optical axis direction directed in the normal direction of an image sensor 8 (described later) of the analog / digital video signal processing main line 6. The imaging lens 5 collects light from the optical axis direction on the image sensor 8.
The analog / digital video signal processing main line 6 includes an image sensor 8, a PGA (Programmable Gain Amplifier) 9, and an ADC (Analog-to-Digital Converter) 10.

The image sensor 8 converts the light collected by the imaging lens 5 into electric power. Further, the image sensor 8 receives the shutter speed (exposure time) corresponding to the shutter speed setting command (described later) output from the exposure control unit 7 (shutter speed / gain setting unit 13 (described later)). An electronic shutter function for outputting the power converted into the exposure time to the PGA 9 is provided.
The PGA 9 amplifies the power output from the image sensor 8 with a gain (amplification factor) corresponding to a gain setting command (described later) output from the exposure control unit 7 (shutter speed / gain setting unit 13). Then, the PGA 9 outputs the amplified power to the ADC 10.

Further, the ADC 10 digitizes the power output from the PGA 9 to generate a main line video signal (image data including information indicating the luminance value of each pixel). Then, the ADC 9 outputs the generated main line video signal to the exposure control unit 7 (luminance photometry unit 11 (described later), exposure error detector 12 (described later)) and the wide D range image generation unit 3.
The exposure control unit 7 includes a luminance photometry unit 11, an exposure error detector 12, and a shutter speed / gain setting unit 13.

The luminance metering unit 11 integrates the luminance value of each pixel included in the image of one screen (one frame, one field) indicated by the main video signal output from the ADC 10 of the analog / digital video signal processing main 6. The integration result is output to the exposure error detector 12 as a luminance evaluation value.
In the present embodiment, the luminance evaluation value is generated by simply integrating the luminance value of each pixel included in the image for one screen. However, the present invention is not limited to this. For example, assuming that the subject is imaged near the center, the pixels near the center may be integrated by weighting, or the image may be divided into a plurality of regions and integrated by weighting each region. It may be.

The exposure error detector 12 compares the brightness evaluation value output from the brightness photometry unit 11 with a brightness reference value (exposure target value) set by an external microcontroller (not shown). That is, the exposure error detector 12 determines the magnitude relationship between the luminance evaluation value and the exposure target value, and outputs the determination result to the shutter speed / gain setting unit 13.
Further, when the determination result output from the exposure error detector 12 indicates that the brightness evaluation value is larger than the exposure target value, the shutter speed / gain setting unit 13 sets the shutter speed (exposure time) until then. Command (hereinafter also referred to as a “shutter speed setting command” or “gain setting command”) for setting a shorter time or a smaller gain is output to the analog / digital video signal processing main line 6 (image sensor 8, PGA 9). To do. Further, when the determination result output from the exposure error detector 12 indicates that the brightness evaluation value is smaller than the exposure target value, the shutter speed / gain setting unit 13 sets the shutter speed longer than before. And a command for setting a large gain are output to the analog / digital video signal processing main line 6. At that time, the shutter speed T is set to an integer N times the shortest exposure time Tmin determined by the structure of the image sensor 8, and the gain M is set to an integer value.

In the present embodiment, an example in which the shutter speed and the gain are switched and set in units of frames is shown, but the present invention is not limited to this. For example, it may be set by switching in units of a pixel area composed of a plurality of pixels forming an image, in units of pixel columns, or in units of pixels.
The shutter speed / gain setting unit 13 is based on the shutter speed T and gain M indicated by the shutter speed setting command and gain setting command output to the analog / digital video signal processing main line 6 according to the following equation (1). An imaging condition value V is generated. Then, the shutter speed / gain setting unit 13 outputs the generated imaging condition value to the wide D range image generation unit 3.

V = (Tmax × Mmax) / (T × M)
= ((Tmin × Nmax) × Mmax) / ((Tmin × N) × M)
= (Nmax × Mmax) / (N × M) (1)

However, Tmax is the longest exposure time allowed by the image sensor 8, Mmax is the maximum gain allowed by the PGA 9, and Nmax (= Tmax / Tmin) is an integer.
In the present embodiment, an example in which the imaging condition value V is switched and set in units of frames has been described. However, the present invention is not limited to this. For example, it may be set by switching in units of a pixel area composed of a plurality of pixels forming an image, in units of pixel columns, or in units of pixels.

  Further, as shown in FIG. 1, the wide D range image generation unit 3 is connected to the main line video signal I (each image for one screen) sequentially output from the imaging unit 2 (ADC 10 of the analog / digital video signal processing main line 6). ) Is sequentially multiplied by the imaging condition value V output from the imaging unit 2 (shutter speed / gain setting unit 13 of the exposure control unit 7) to generate a wide D range image signal WI. Then, the wide D range image generation unit 3 sequentially outputs the generated wide D range image signal WI to the motion detection unit 4 (memory arbiter 15 (described later), difference unit 17 (described later)).

Further, as shown in FIG. 3, the motion detection unit 4 includes a pixel counter / control unit 14, a memory arbiter 15, an image memory 16, a difference unit 17, and a determination unit 18.
The pixel counter / control unit 14 detects various synchronization signals (horizontal synchronization signal, vertical synchronization signal, pixel synchronization signal, etc.) in the imaging unit 2, and based on the detected synchronization signal, the wide D range image generation unit 3 The XY coordinate position of the pixel indicated by the wide D range image signal WI output from is calculated. The pixel counter / control unit 14 sequentially inputs the wide D range image signal WI, and outputs a predetermined control signal to the memory arbiter 15 every time the XY coordinate position changes, and also corresponds to the XY coordinate position. An address / control signal indicating an address for recording the luminance value of the pixel is output to the image memory 16.

  Further, when a control signal is output from the pixel counter / control unit 14, the memory arbiter 15 has an address corresponding to the XY coordinate position of the pixel indicated by the wide D range signal WI output from the wide D range image generation unit 3. The information of the previous frame (information on the luminance value of the corresponding pixel) is read out from. Then, the memory arbiter 15 outputs the read information to the difference unit 17 and writes information on the luminance value indicated by the wide D range image signal WI to the corresponding address.

  Further, when there is a read request from the memory arbiter 15, the image memory 16 stores the information of the previous frame (information of the luminance value of the corresponding pixel) from the address indicated by the address / control signal output from the pixel counter / control unit 14. The arbiter 15 is read out. When there is a write request from the memory arbiter 15, information on the luminance value corresponding to the write request is written to the address indicated by the address / control signal output from the pixel counter / control unit 14.

The difference unit 17 calculates a difference between the luminance value indicated by the information output from the memory arbiter 15 and the luminance value of the pixel indicated by the wide D range image signal output from the wide D range image generation unit 3. . Then, the difference unit 14 outputs the calculation result to the determination unit 18.
Further, the determination unit 18 detects the movement of the subject based on the magnitude of the calculation result output from the difference unit 17, and transfers the detection result to a CPU or a motion detection memory (not shown).

<Specific operation of motion detection device>
Next, the operation of the motion detection apparatus according to the present invention will be described based on a specific situation. In this example, the imaging condition is updated in units of frames.

Now, an imaging operation is performed without any movement, and the imaging condition value at this time is set to V1. A wide D range image signal is generated by the imaging condition value V1 (described later).
Then, it is assumed that the luminance values of some pixels in the current frame image indicated by the main line video signal increase due to the movement of the subject during the execution of the subject motion detection. Then, as shown in FIG. 2, the luminance metering unit 11 calculates the luminance evaluation value sufficiently larger than the exposure target value, and the exposure error detector 12 gives a determination result indicating that to the shutter speed / gain setting unit 13. Is output. Then, the shutter speed / gain setting unit 13 outputs a shutter speed setting command to the image sensor 8 of the analog / digital video signal processing main line 6, outputs a gain setting command to the PGA 9, and the imaging condition value V2 is wide D. It is output to the range image generation unit 3. The above operation is executed immediately after the imaging operation for the current frame is completed, and in the next frame, the imaging operation is performed under new imaging conditions.

  When the detection of the next frame is started, the image sensor 8 converts the light collected by the imaging lens 5 into electric power at the shutter speed (exposure time) corresponding to the shutter speed setting command, and the conversion. The generated power is output to the PGA 9. The PGA 9 amplifies the power output from the image sensor 8 with a gain (amplification factor) corresponding to the gain setting command, and outputs the amplified power to the ADC 10. The output power is digitized by the ADC 10 to generate a main line video signal, and the main line video signal is output to the wide D range image generation unit 3, the luminance photometry unit 11, and the exposure error detector 12. .

  Then, as shown in FIG. 1, the imaging conditions output from the shutter speed / gain setting unit 13 of the exposure control unit 7 of the imaging unit 2 to the main video signal I output by the wide D range image generation unit 3. The wide D range image signal WI is generated by sequentially multiplying the value V2. That is, the wide D range in which the main video signal I is normalized by the shutter speed T and gain M used when the main video signal I is generated, and the luminance value of the background image is kept constant in all frames. An image signal WI is generated. Then, the wide D range image generation unit 3 sequentially outputs the generated wide D range image signal WI to the memory arbiter 15 and the difference unit 17 of the motion detection unit 4.

  At the same time, as shown in FIG. 3, the pixel counter / control unit 14 detects various synchronization signals (horizontal synchronization signal, vertical synchronization signal, pixel synchronization signal, etc.) in the imaging unit 2, and the detected synchronization signal Based on this, the XY coordinate position of the pixel indicated by the wide D range image signal WI output from the wide D range image generation unit 3 is detected. Each time the detected XY coordinate position is changed by the pixel counter / control unit 14, a control signal is output to the memory arbiter 15, and an address / control signal is output to the image memory 16. Further, each time the control signal is output, the memory arbiter 15 reads the previous frame information (luminance value of the corresponding pixel) from the address corresponding to the XY coordinates of the pixel indicated by the wide D range image signal WI. Then, the memory arbiter 15 outputs the read information to the difference unit 17 and stores information on the luminance value indicated by the wide D range image signal WI at the corresponding address. Further, the difference between the luminance value indicated by the output information and the luminance value of the pixel indicated by the wide D range image signal WI is greatly calculated by the difference unit 17 in the portion where the subject is moving, and the background portion. Is almost calculated as “0”, and the calculation result is output to the determination unit 18. Then, the determination unit 18 appropriately detects only a region where the subject moves based on the size of the output calculation result, and transfers the detection result to a CPU or a motion detection memory (not shown).

  As described above, according to the motion detection device of the present embodiment, when the shutter speed and the gain of the electronic shutter change due to the exposure control operation, the luminance value of the image is normalized by the shutter speed and the gain. The brightness value of the image can be kept constant. Therefore, it is possible to prevent a frame difference from occurring in the background image. For example, it is possible to improve the accuracy of motion detection compared to a method of detecting the motion of the subject using the image captured by the imaging unit 2 as it is. it can.

As described above, the wide D range image generation unit 3 in FIG. 1 constitutes the normalization means described in the claims, and similarly, the motion detection unit 4 in FIG. 1 and the difference unit 17 in FIG. 3 constitute the detection means. .
The wide D range image generation unit 3 in FIG. 1 constitutes normalization means described in the claims. Further, the motion detection device of the present invention is not limited to the contents of the above embodiment, and can be appropriately changed without departing from the spirit of the present invention.

It is a block diagram which shows schematic structure of the motion detection means of this invention. It is a block diagram which shows the internal structure of the imaging part of FIG. It is a block diagram which shows the internal structure of the motion detection part of FIG.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Detection apparatus, 2 ... Imaging part, 3 ... Range image generation part, 4 ... Detection part, 5 ... Imaging lens, 6 ... Analog / digital image signal processing main line, 7 ... Exposure control part, 8 ... Image sensor, 9 ... PGA, 10 ... ADC, 11 ... luminance metering unit, 12 ... exposure error detector, 13 ... shutter speed / gain setting unit, 14 ... pixel counter / control unit, 15 ... memory arbiter, 16 ... image memory, 17 ... difference unit , 18 ... determination unit.

Claims (2)

  A motion detection device that detects a motion of a subject by detecting a portion in which a luminance value has changed between a plurality of images captured at different times, the shutter speed of an electronic shutter at the time of capturing the image, and the image The normalization means for normalizing the luminance value of the image at least one of the gains multiplied by the output of the image sensor that picked up the image, and the luminance value changed between the images whose luminance values were normalized by the normalization means A motion detection apparatus comprising: a detecting means for detecting a portion.   The normalization unit multiplies the luminance value of the image by a conversion coefficient calculated by dividing a product of a maximum shutter speed and a maximum gain by the shutter speed and the gain. The motion detection apparatus described in 1.

Images