JP2016127365A - Image processing apparatus, image processing method, imaging system, image processing system and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow for control of the intensity of a reflection component regardless of the polarization state of the reflection component in an optical member, when acquiring the image of a subject through the optical member having optical transparency and light reflectivity.SOLUTION: The mixed light of the light from the inside of a vehicle transmitted through a front glass and the light reflected on the front glass enters an imaging unit 101. A polarization image generation unit 103 and a luminance image generation unit 102 generate a polarization image and a luminance image, respectively. A reflection component estimation unit 104 estimates the reflection light component in the front glass, from the output of the luminance image generation unit 102 and the output of the polarization image generation unit 103. The detail of the reflection component estimation unit 104 is described below. A reflection component removal unit 105 generates the visible image of the inside of a vehicle, by removing the estimate value of reflection light component from the luminance image.SELECTED DRAWING: Figure 8

Description

  The present invention relates to an image processing apparatus, an image processing method, an imaging system, an image processing system, and a program.

  Conventionally, there is a vehicle traffic monitoring system for the purpose of monitoring a passing vehicle on a general road or a highway. As a known traffic monitoring system, a system that mainly acquires a person such as a vehicle number or a driver in a vehicle using a monitoring camera is known.

  That is, for example, in Patent Document 1, in order to identify the driver in the vehicle, when acquiring an image of the driver as a subject through the windshield of the vehicle (through the windshield), for example, light irradiated by sunlight Of the above components, the reflection component from the windshield and the reflection component from the driver's face are separated using a polarizing filter and the reflection component from the windshield is removed. A system for acquiring the images of is described.

  However, in the system described in Patent Document 1, the performance of removing the reflection component on the windshield depends on the polarization state of the reflection component. That is, there is a problem that the reflection component cannot be removed when the reflection component is not polarized or when the polarization direction of the reflection component is the same as the polarization direction of the polarization filter.

This will be described with reference to FIG.
FIG. 13A shows the relationship between the reflected light from the windshield (with polarization), the transmitted light from the windshield (without polarization), and the detection light from the camera when no polarizing filter is provided. In this case, the intensity of the detection light is the sum of the intensity of the mixed light of the reflected light and the transmitted light, that is, the sum of the two kinds of light.

  In FIG. 13B, by providing a polarizing filter having a specific polarization axis, the reflected light on the windshield, the transmitted light on the windshield, and the camera in an ideal case where the reflected light on the windshield is completely removed. The relationship of the detected light is shown. In this case, the intensity of the detection light is ideally ½ of the intensity of the transmitted light.

  FIG. 13C shows the case where a part of the reflected light from the windshield is removed even if a polarizing filter having a specific polarization axis is provided, the reflected light from the windshield, the transmitted light from the windshield, and the detected light from the camera. Shows the relationship. In this case, the intensity of the detected light is the sum of 1/2 of the intensity of the transmitted light (ideal value) and the intensity of the reflected light that has not been removed (residual).

  The present invention has been made in order to solve such a problem, and its purpose is to obtain an image of a subject through an optical member having light transmittance and light reflection such as a windshield of a vehicle. The intensity of the reflection component can be controlled regardless of the polarization state of the reflection component at the optical member.

  The image processing apparatus of the present invention captures image information corresponding to different polarization states generated by imaging a subject through an optical member having light transmittance and light reflectivity by an imaging device including polarization members having different polarization states. An image processing device for processing, wherein image processing means for controlling a component corresponding to reflected light from the optical member in image information corresponding to the different polarization state using image information corresponding to the different polarization state An image processing apparatus.

  According to the present invention, when an image of a subject is acquired through an optical member having light transparency and light reflection, the intensity of the reflection component can be controlled regardless of the polarization state of the reflection component on the optical member. .

It is a figure for demonstrating the installation environment of the image processing system which concerns on embodiment of this invention. It is a figure for demonstrating typically the path | route of the light ray which injects into the image processing system which concerns on embodiment of this invention. 1 is a diagram illustrating a schematic configuration of an information processing system according to an embodiment of the present invention. It is a block diagram which shows the hardware constitutions of the polarization camera which concerns on embodiment of this invention. It is a block diagram which shows the hardware constitutions of the video server which concerns on embodiment of this invention. It is a figure which shows the structural example of the image sensor in FIG. It is a figure which shows the example of the correspondence of the polarizing filter and color filter in FIG. It is a block diagram which shows the structure of the polarization camera which concerns on the 1st Embodiment of this invention. It is a flowchart which shows the process of the reflection component estimation part in FIG. It is a figure for demonstrating the input information of the reflection component estimation part in FIG. It is a block diagram which shows the structure of the polarization camera which concerns on the 2nd Embodiment of this invention. It is a flowchart which shows schematic operation | movement of the information processing system which concerns on embodiment of this invention. It is a figure for demonstrating the problem in the case of removing the reflective component in a windshield with a polarizing filter.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
<Installation environment of image processing system>
FIG. 1 is a diagram for explaining an installation environment of an image processing system according to an embodiment of the present invention, and FIG. 2 is a schematic view of a path of light rays incident on the image processing system according to the embodiment of the present invention. It is a figure for demonstrating.

  As shown in FIG. 1, the first image processing system 1-1 to the n-th (n is an integer of 2 or more) image processing systems 1-n (the image processing systems 1 are referred to when the image processing systems are not distinguished). Is installed on the side road 3. Each image processing system 1-1 to 1-n is housed in a case with a window called a cabin held at a predetermined height.

  As shown in FIG. 2, the image processing system 1 includes light (hereinafter referred to as sunlight windshield) that is emitted from the sun and reflected by the windshield of the vehicle 4 (an optical member having light transmittance and light reflectivity). Light reflected from the sun, transmitted through the windshield of the vehicle 4, reflected by the driver (subject) inside the vehicle 4, and again transmitted through the windshield (hereinafter referred to as the sunlight target portion). Reflected light) and light reflected from the windshield of the vehicle 4 (hereinafter referred to as reflected light of the scattered light) enters the vehicle 4 and is scattered by clouds such as mottled clouds.

  In addition, although illustration was abbreviate | omitted, as reflected light, reflected light such as scattered light by a building existing in the vicinity or sunlight from a tree may enter as reflected light. Hereinafter, such reflected light is referred to as reflected light from the environment.

  Here, the reflected light of the sunlight on the windshield is light having a large amount of polarization components in a certain direction, as indicated by the vertical arrows. The reflected light of the scattered light from the cloud is also light with many polarization components in the same direction. Moreover, the reflected light from the sunlight target portion is non-polarized light as indicated by the vertical arrows and the horizontal arrows.

  The image processing system 1 passes incident light through a predetermined polarizing filter, converts the incident light into an electrical signal, and performs predetermined image processing to be described later, thereby reflecting light reflected by the sunlight reflected on the windshield and the environment. Component is reduced, and a visualized image inside the vehicle 4 is generated and transmitted to the data center via the network.

<Schematic configuration of information processing system>
FIG. 3 is a diagram showing a schematic configuration of the information processing system according to the embodiment of the present invention.
As shown, the information processing system includes a first image processing system 1-1, a second image processing system 1-2,... N-th image processing system 1-n, and each image processing system. The data center 6 is connected via a network 5 such as the Internet.

  The first image processing system 1-1 includes a polarization camera 100-1 and a video server 200-1. Polarization camera 100-1 passes incident light through a predetermined polarizing filter, converts it into an electrical signal, and performs predetermined image processing to be described later, thereby reflecting the reflected light component of sunlight on the windshield and the environment. The light component is reduced, and a visualized image inside the vehicle 4 is generated and output to the video server 200-1. The video server 200-1 stores the visualized image output from the polarization camera 100-1, and then transmits it to the data center 6. The data center 6 receives the visualized image transmitted from the video server 200-1 and performs recognition processing and the like.

  The second image processing system 1-2 to the n-th image processing system 1-n are similarly configured and operate in the same manner. In the following description, the polarization camera 100 and the video server 200 are used when the image processing system is not distinguished from the polarization camera and the video server.

<Hardware configuration of polarization camera>
FIG. 4 is a block diagram illustrating a hardware configuration of the polarization camera.
As illustrated, the polarization camera 100 includes an imaging unit 101 as an imaging device and a signal processing board 170.

  The imaging unit 101 includes an image sensor 150 and a sensor controller 160. The image sensor 150 converts incident light into an electrical signal and outputs it to the sensor controller 160. The sensor controller 160 plays a role of exposure control, readout control of the image sensor 150, communication with the signal processing board 170, and transmission of image data.

  The signal processing board 170 includes an FPGA 171, a CPU 172, a ROM (including a rewritable one) 173, a RAM 174, a serial I / F (interface) 175, a data bus 176, and a serial bus 177.

  Image data output from the sensor controller 160 is transferred to the RAM 174 via the data bus 176. The FPGA 171 executes processing that requires real-time processing on the image data stored in the RAM 174 and writes it back to the RAM 174. The CPU 172 controls the operation of the entire signal processing board 170, performs image processing, and the like by executing a computer program stored in the ROM 173 using the RAM 174 as a work area. In addition, the CPU 172 transmits the image data that has undergone the image processing (here, a visualized image inside the vehicle) to the video server 200 via the serial I / F 175. Further, the FPGA 171 or the CPU 172 transmits and receives various setting data such as changing the exposure control value of the image sensor 150 and changing the image reading parameter.

<Hardware configuration of video server>
FIG. 5 is a block diagram illustrating a hardware configuration of the video server.
As illustrated, the video server 200 includes a bus 210, a CPU 220 connected to the bus 210, a ROM 230, a RAM 240, an HDD (hard disk) 250, a serial I / F 260, and a network I / F 270.

  The CPU 220 controls the overall operation of the video server 200 by executing a computer program stored in the ROM 230 using the RAM 240 as a work area. The serial I / F 260 receives the image data (visualized image inside the vehicle) transmitted from the serial I / F 175 of the imaging unit 101. The HDD 250 stores image data received by the serial I / F 260. The network I / F 270 communicates with the data center 6 via the network 5 to transmit a visualized image inside the vehicle to the data center 6.

<Configuration of image sensor>
FIG. 6 is a diagram illustrating a configuration example of the image sensor in FIG. Here, FIG. 6A is a configuration example in which a polarizing filter layer as a polarizing member is disposed in front of the lens, and FIG. 6B is a configuration example in which the polarizing filter layer is disposed behind the lens.

  In the image sensor shown in FIG. 6A, a polarizing filter layer 151, a microlens 152, a color filter layer 153, and a light receiving element 154 are arranged in this order. In the image sensor shown in FIG. 6B, the microlens 152, the polarizing filter layer 151, the color filter layer 153, and the light receiving element 154 are arranged in this order.

  The color filters constituting the color filter layer 153 have different colors for individual pixels. Generally, it consists of three types of R (red), G (green), and B (blue), and a single matrix is formed by 4 pixels consisting of 2 pixels of G and 1 pixel each of R and B. The matrix is periodically arranged in the horizontal direction and the vertical direction.

  In the polarizing filter layer 151, the pixel pattern of the polarizing filter is set in accordance with the pixel pattern of the color filter. The correspondence between the pixel patterns will be described later.

  Note that the position of the polarizing filter layer 151 is not limited to the position shown in FIGS. 6A and 6B, and may be arranged anywhere as long as it is set to function individually for each pixel.

<Correspondence between polarization filter and color filter>
FIG. 7 is a diagram illustrating an example of a correspondence relationship between the pixel pattern of the polarization filter layer and the pixel pattern of the color filter layer in FIG. Here, six examples shown in FIGS. 7A to 7F are given as examples of the correspondence relationship between the single matrix of the polarization filter and the single matrix of the color filter. Hereinafter, the upper left, upper right, lower left, and lower right pixels of the matrix in these drawings will be described as Pix1, Pix2, Pix3, and Pix4.

  In the first example shown in FIG. 7A, R (red filter pixel), G (green filter pixel), G, B (blue filter pixel) of the color filter layer 153 for Pix1, Pix2, Pix3, and Pix4; C, C, V, and C of the polarizing filter layer 151 are associated with each other. Here, C (Clear) is a pixel that transmits all polarization components (a pixel that does not include a polarization filter). Further, V (Vertical) is a pixel that transmits a vertical electric field component (a pixel having a polarization filter of a specific direction axis). H (Horizontal), which will be described later, is a pixel that transmits a horizontal electric field component (a pixel having a polarization filter with a specific direction axis).

In the pixel arrangement shown in this figure, luminance color image information is used at Pix1, Pix2, and Pix4. In addition, a polarization image is generated from Pix2 and Pix3. That is, for example, Pix3 (G pixel and V pixel) can be used as it is, or a difference pixel between Pix2 and Pix3 can be used. The following equation [1] is established as the difference signal.
“Luminance value of (G pixel and H pixel)” = “Luminance value of Pix2 (G pixel and C pixel)” − “Luminance value of Pix3 (G pixel and V pixel)” Equation [1]

  Here, the luminance color image is an image calculated and generated from R, G, and B3 color components. A polarization image is an image generated by a pixel that transmits a vertical or horizontal electric field vibration component (has a polarization axis). In addition, when outputting a luminance color image and a polarization image, in addition to the above calculation, an interpolation process using adjacent matrix pixels of a single matrix is also performed.

  In the second example shown in FIG. 7B, R, G, G, and B of the color filter layer 153 and C, C, H, and C of the polarization filter layer 151 are associated with Pix1, Pix2, Pix3, and Pix4. It has been. That is, V in the first example is changed to H. Since the above-described difference formula is established, a polarization image can be obtained regardless of whether H pixels or V pixels are arranged.

  In the third example shown in FIG. 7C, R, G, G, and B of the color filter layer 153 and C, V, H, and C of the polarization filter layer 151 are associated with Pix1, Pix2, Pix3, and Pix4. It has been. Polarized images are generated from Pix2 and Pix3. For the luminance color image, an R component and a B component are generated from Pix1 and Pix4, respectively, and a G component is generated from “Pix2 + Pix3”. This is based on the idea that the difference (H = C−V, that is, C = V + H) in the above equation is established.

  In the fourth example shown in FIG. 7D, R, G, G, and B of the color filter layer 153 and C, H, V, and C of the polarization filter layer 151 are associated with Pix1, Pix2, Pix3, and Pix4. It has been.

  In the third and fourth examples, the luminance component of the entire image is made uniform, and the dynamic range is improved. Although depending on the subject, the G pixel generally has higher brightness than the R pixel and the B pixel, and is likely to be saturated depending on the shooting conditions. Similarly, when the polarizing filters V and H are combined with the G pixel, the luminance is approximately twice as high when the polarizing filter C is combined. In the first example and the second example, of the four pixels, Pix2 The G pixel is in a higher luminance state than the surrounding pixels. On the other hand, by using the patterns of the third example and the fourth example, it is possible to make the luminance uniform among the four pixels constituting the single matrix.

  In the fifth example shown in FIG. 7E, R, G, G, and B of the color filter layer 153 and V, V, H, and V of the polarizing filter layer 151 correspond to Pix1, Pix2, Pix3, and Pix4. It has been. In the sixth example shown in FIG. 7F, R, G, G, and B of the color filter layer 153 and H, H, V, and H of the polarizing filter layer 151 are set for Pix1, Pix2, Pix3, and Pix4. It is associated.

  In general, a pixel provided with a polarizing filter has lower luminance than surrounding pixels. For example, in the first example, the brightness of Pix1, 2, and 4 of the surrounding pixels is higher than Pix3 used for generating the polarization image.

  In this pattern configuration, more light components are mixed from Pix1, 2, 4 with high luminance to Pix3 with low luminance, causing crosstalk, and Pix3 has more error components than the signal value that Pix3 should originally take become. This phenomenon is particularly prominent when the polarizing filter layer 151 as shown in FIG. 6A is disposed farther from the light receiving element 154.

  In order to suppress this, it is effective to adopt a configuration in which the luminance values of the surrounding pixels Pix1, 2, and 4 are relatively lowered with respect to the polarization pixel Pix3. This effect can be obtained by installing the polarizing filter V or H for Pix1, 2, 4 as shown in the fifth example and the sixth example.

<Polarized camera according to the first embodiment>
FIG. 8 is a block diagram showing the configuration of the polarization camera according to the first embodiment of the present invention.
The polarization camera includes an imaging unit 101, a luminance image generation unit 102 as a first unit, a polarization image generation unit 103, a reflection component estimation unit 104 as a second unit and a third unit, and a fourth unit. The reflection component removal processing unit 105 is provided.

  Here, the imaging unit 101 is the imaging unit 101 in FIG. The other parts constitute the image processing apparatus. This image processing apparatus includes an FPGA 171, a CPU 172, a ROM 173, and a RAM 174.

  Light that has passed through the windshield from the inside of the vehicle (transmitted light) and reflected light from the windshield enters the imaging unit 101. The luminance image generation unit 102 and the polarization image generation unit 103 generate a luminance image and a polarization image, respectively, by the calculation described with reference to FIG.

  The reflection component estimation unit 104 estimates a reflected light component on the windshield based on the luminance image generated by the luminance image generation unit 102 and the polarization image generated by the polarization image generation unit 103. Details of the reflection component estimation unit 104 will be described later. The reflection component removal processing unit 105 generates a visualized image inside the vehicle by removing the estimated value of the reflected light component from the luminance image.

<Processing of reflection component estimation unit>
FIG. 9 is a flowchart showing processing of the reflection component estimation unit in FIG. 8, and FIG. 10 is a diagram for explaining input information of the reflection component estimation unit in FIG.

  As shown in FIG. 9, first, a statistic of the luminance value I of the luminance image generated by the luminance image generation unit 102 is calculated (step S1). Examples of the statistics to be calculated include an average value and a standard deviation value in the entire screen or a predetermined area of the screen.

Next, a difference value between the luminance value I as the first information and K ₁ * P which is a value obtained by multiplying the pixel value P of the polarization image generated by the polarization image generation unit 103 in advance and the coefficient K ₁ A statistic of Δ (= I−K ₁ * P) is calculated (step S2).

Here, the luminance value I and the pixel value P of the polarization image will be described with reference to FIG.
FIG. 10A shows the luminance value of the C pixel that receives the reflected light (with polarization) with intensity a and the transmitted light with intensity b (without polarization). In this case, the pixel value of the C pixel corresponding to the luminance value I is “a + b”.

FIG. 10B shows the pixel values of the V pixel and the H pixel that receive the reflected light (with polarization) with intensity a and the transmitted light with intensity b (without polarization). In this case, the pixel values of the V pixel and the H pixel corresponding to the pixel value P are “a / K ₀ + b / K ₁ ”. Here, K ₀ and K ₁ are coefficients, K ₁ is ideally 2, and K ₀ varies depending on the ratio of polarization components other than the vertical direction or the horizontal direction with respect to the entire reflected light (with polarization). In the present embodiment, K ₁ is a value of “I / P” when non-polarized light such as natural light is incident on the imaging unit 101 in advance (when a = 0 in FIGS. 10A and 10B). _{1 is} assumed.

10A and 10B, the following formula [2] is obtained, and it can be seen that the value of Δ is proportional to the intensity a of the reflected light. Δ corresponds to the second information according to the present invention.
Δ (= I−K ₁ * P) = {(a + b) −K ₁ (a / K ₀ + b / K ₁ )} = {1− (K ₁ / K 0)} a (2)

Next, a coefficient K ₂ is obtained from the statistic of the luminance value I calculated in step S 1 and the statistic of Δ calculated in step S ₂ (step S 3), and the coefficient K ₂ is multiplied by Δ to obtain a reflection component. Intensity K ₂ * Δ is calculated (step S4). K ₂ * Δ corresponds to control information and third information in the present invention.

Here, the coefficient K ₂ is calculated as “K ₂ = statistic of I / statistic of Δ”. In this case as the statistic employed, the value of K ₂ is selected to be less than "average value of the average value / delta of I". That is, when Δ is an average value, I is set to a value smaller than the average value (for example, “average value−3σ” (σ is a standard deviation value)). The reason will be explained. Assuming that the signal is normally distributed by taking the average value ± 3σ of the I signal, 99.7% of the signals are included. That is, the process “I−K ₂ * Δ” in the reflection component removal processing unit 105 is based on K ₂ obtained by dividing the average value −3σ of the I signal by the lower limit value of the signal and the average value −3σ by the average value of Δ. By performing the above, information in which the component of reflected light is reduced can be obtained from most luminance images. When the value of K ₂ is “average value of I / average value of Δ”, the value of K ₂ is larger than “average value of a / average value of Δ”, so that “I−K ₂ * Due to “Δ”, information of the low-luminance portion of the luminance image is lost.

  As described above, according to the polarization camera of the first embodiment of the present invention, information indicating the intensity of the reflection component is acquired and the reflection component is removed regardless of the polarization state of the reflection component on the vehicle windshield. can do.

<Polarized camera according to the second embodiment>
FIG. 11 is a block diagram showing a configuration of a polarization camera according to the second embodiment of the present invention. In this figure, the same reference numerals as those in FIG. 8 are given to the same or corresponding parts as those in FIG. 8 (first embodiment). Description of these portions is omitted unless necessary.

In this polarization camera, as a fourth means according to the present invention, a reflection component enhancement processing unit 106 is provided instead of the reflection component removal processing unit 105 of the polarization camera according to the first embodiment. The reflection component enhancement processing unit 106 increases the reflection component by adding the reflection component estimated value K ₂ * Δ to the luminance value I of the luminance image.

  As a result, the output of the reflection component enhancement processing unit 106 becomes an invisible image in which the subject in the windshield is difficult to see. Therefore, this polarization camera protects the privacy of the imaged person (for example, the data center 6 does not use the image acquired by this polarization camera to identify the driver, and separately uses the first embodiment). This is useful when providing a visible image acquired by such a camera to the police in order to identify the driver, or when making it impossible to identify an internal person when imaging a restaurant or the like from outside.

<Outline of information processing system>
FIG. 12 is a flowchart showing a schematic operation of the information processing system according to the embodiment of the present invention.

  In the polarization camera 100, the CPU 172 determines whether or not the vehicle is shooting from the image from the imaging unit 101 (step S11). This determination is made based on, for example, a change in the background image of the previous frame.

  Monitoring is continued while the vehicle is not photographed (step S11: NO → step S11). If the vehicle is photographed (step S1: YES), image processing is started (step S12). This image processing may be either image processing in the polarization camera (FIG. 8 or FIG. 10) according to the first or second embodiment described above.

  Next, the polarization camera 100 outputs the visualized image generated by the image processing to the video server 200 (Step S13), and returns to the determination processing whether or not the vehicle is photographed (Step S11). Thereafter, the polarization camera 100 repeats steps S11 to S13.

  The video server 200 stores the visualized image in the HDD 250 as an image storage unit (step S14), and transmits the visualized image to the data center 6 (step S15). The data center 6 receives the visualized image and confirms the image (step S16).

In addition, this invention is not limited to said embodiment, For example, a deformation | transformation like the following (1) thru | or (6) is possible.
(1) The average value K ₃ and standard deviation value σ of “I value / Δ value” are calculated, and “K ₃ -3σ” is set as the coefficient K ₂ .
(2) A value of Δ for the entire screen is calculated, a histogram thereof is obtained, and pixels having a frequency exceeding a certain value are extracted, thereby extracting a plurality of regions having high reflected light intensity on the screen. Then, a reflected light component reduction process is performed on each of these regions, K2 optimum for each region is calculated, and the reflected light component is reduced. Thereby, the visualization image inside a several vehicle currently imaged simultaneously with one polarization camera can be acquired simultaneously.
(3) “H (Horizontal) + V (Vertical)” is used in place of “I” in FIGS. 8 and 11, and “H” or “V” is used in place of “P”. It can be applied not only to the polarization signal of the above, but also to the signal obtained by the polarization filters in the horizontal and vertical directions.
(4) A configuration in which the reflection component removal processing and the reflection component enhancement processing can be switched or selected according to the application.
(5) Two polarization cameras are prepared to provide a stereo camera having a predetermined baseline length. In the stereo camera, which is realized by a polarization camera, a distance image can be generated in addition to the image processing. Since it is possible to acquire a distance image from the image processing system to the vehicle where the vehicle is imaged, it is possible to flexibly cope with the setting of the timing for starting the image processing (step S2 in FIG. 12).
(6) The signal processing board 170 of the polarization camera 100 is provided in a controller separate from the polarization camera 100, and an imaging system in which the polarization camera and the controller cooperate is provided.

  DESCRIPTION OF SYMBOLS 1,1-1 thru | or 1-n ... Image processing system, 6 ... Data center, 100, 100-1 ... Polarized camera, 101 ... Imaging part, 102 ... Luminance image generation part, 103 ... Polarized image generation part, 104 ... Reflection Component estimation unit, 105 ... reflection component removal processing unit, 106 ... reflection component enhancement processing unit, 200, 200-1 ... video server.

JP 2011-165005 A

Claims (10)

An image processing apparatus that processes image information corresponding to different polarization states generated by imaging a subject through an optical member having light transmittance and light reflectivity by an imaging device including polarization members having different polarization states. ,
An image processing apparatus comprising image processing means for controlling a component corresponding to reflected light from the optical member in the image information corresponding to the different polarization states, using image information corresponding to the different polarization states. The image processing apparatus according to claim 1,
The image processing device, wherein the image processing means reduces a component corresponding to the reflected light. The image processing apparatus according to claim 1,
The image processing device increases the component corresponding to the reflected light. The image processing apparatus according to any one of claims 1 to 3,
The image processing means includes: first means for obtaining first information representing intensity of mixed light of transmitted light and reflected light at the optical member; and the reflected light at the optical member from the image information. A second means for obtaining second information proportional to the intensity; a third means for obtaining an estimated value of the intensity of the reflected light from the first information and the second information; and And a fourth means for controlling the value of the first information based on the image processing apparatus. The image processing apparatus according to claim 4,
The image processing apparatus, wherein the third means calculates the coefficient based on an average value of the first information value and an average value of the second information value in a predetermined region of the image. In the image processing device according to claim 4 or 5,
The different polarization states are unpolarized and polarized,
The second information is a value of a ratio between a pixel value of a non-polarized pixel and a pixel value of a polarized image when the imaging apparatus receives non-polarized light from the pixel value of the non-polarized pixel. An image processing apparatus, which is a value obtained by subtracting a value obtained by multiplying a coefficient by a pixel value of a pixel with polarization. An image processing method for processing image information corresponding to different polarization states generated by imaging a subject through an optical member having light transmittance and light reflectivity by an imaging device including polarization members having different polarization states. ,
Generating control information using image information corresponding to the different polarization states;
Using the control information to control a component corresponding to the reflected light from the optical member in the image information corresponding to the different polarization states;
An image processing method. An imaging unit including a polarizing member having different polarization states;
When the imaging unit captures an image of a subject through an optical member having light transmittance and light reflectivity and generates image information corresponding to different polarization states, the image information is used in the optical member in the image information. Image processing means for controlling a component corresponding to the reflected light;
An imaging system. An imaging unit including a polarizing member having different polarization states;
When the imaging unit captures an image of a subject through an optical member having light transmittance and light reflectivity and generates image information corresponding to different polarization states, the image information is used in the optical member in the image information. Image processing means for controlling a component corresponding to the reflected light;
Storage means for storing the output of the image processing means;
An image processing system. A program for causing a computer to function as each unit of the image processing apparatus according to any one of claims 1 to 6.

Images