JP2002071458A - Method for determining light source type - Google Patents

Abstract

PROBLEM TO BE SOLVED: To estimate a type of a light source irradiating a scene and support a treatment for white balance. SOLUTION: The light source is estimated from the intensity difference between rays of visible light and ultraviolet light in the spectra from the scene. If the total intensity of visible light is rather stronger than that of ultraviolet light, the light source is a fluorescent lamp, if the difference is not so large, it is a tungsten lamp and if the ratio of ultraviolet light is rather lager, it is sunlight.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

FIELD OF THE INVENTION The present invention relates generally to digital cameras, and more particularly, to digital cameras having an auxiliary UV (ultraviolet) photosensor. This digital camera is U
The visible light photosensor in the camera along with the V photosensor can be used to determine the type of light source in the scene.

[0002]

BACKGROUND OF THE INVENTION When capturing images with a digital camera, the illumination source of the scene affects the color captured by the camera. In indoor scenes, illumination sources can vary widely, such as tungsten bulbs, halogen lamps, fluorescent lights, sunlight coming through windows, or even xenon light. Each of these types of light sources has a different spectral energy distribution. Light sources of the type that generate light using filaments that glow at high temperatures (eg, tungsten bulbs) are generally characterized by a color temperature defined as a black body that is 50 ° C. higher than the filament of the light source (see FIG. 1). See). The sun can also be characterized as a black body, but exhibits a significantly different distribution from black bodies at such wavelengths because some wavelengths suffer losses due to scattering and absorption in the atmosphere. Due to the various spectral power distributions of the sun, a number of standard spectral power distribution curves have been developed.
One of the standard curves is a D65 corresponding to a color temperature of 6500K.
(See FIG. 2). Clouds in the sky can also affect the spectral distribution of energy reaching the scene from the sun. Time of day also affects the color temperature of the sun (compare noon and sunrise). Color temperature can be affected by whether an object is directly exposed to sunlight or in the shade.

Light sources of the type that excite a phosphor layer to emit fluorescence (eg, fluorescent lamps and xenon lamps) tend to have a spectral distribution specific to the phosphor of the lamp in combination with the mercury vapor spectrum ( See FIG. 3).

Each of these light sources has a different spectral power distribution, which affects the colors captured by the camera in the scene. For example, a white object illuminated by a tungsten bulb will appear yellow in the scene captured by the camera. This is because tungsten bulbs do not produce as much blue light. A white object is an object that reflects a similar amount of red, green, and blue light as strikes the object. When a white object is illuminated by a tungsten bulb, more red light hits the subject than blue light, so more red light is reflected and the camera appears to be yellow. While the human eye adapts to various illuminations to compensate for color shifts, the camera records the actual light of the scene.

[0005] Advantageously, such color shifts introduced by the illumination source can be compensated. This compensation is generally called white balance compensation. For proper white balance compensation, it is necessary to know the light source of the scene. Many methods are currently used to determine the scene light source used for white balance compensation.

One approach looks for the brightest point in the scene and assumes that it is white. The brightest point is then adjusted until it becomes white, after which the balance of the scene is balanced. This method is performed assuming that the brightest points in the scene are from white objects or specular reflections. For example,
There is specular reflection coming from the windshield of the car. Obviously, the brightest points in all scenes are not specular or white objects. Using this method for scenes where the brightest point of the scene is an object of a color other than white can result in significant color shift. In another white balance compensation method, the image is adjusted until the sum of all the regions of the image becomes neutral gray. Both of these methods operate on assumptions about the content of the scene.

In another method, a correlation matrix memory (correlati
On matrix memory), the image data is mapped to color image data under a plurality of different light sources. This method is described in U.S. Patent No. 6,038,339. When using this method, the image data is
All possible light sources need to be mapped to color data. Mapping the image data to each of the possible light sources is a computationally performed process. If the set of possible light sources could be restricted to a particular type of light source (eg daylight), the amount of computation, and thus the time required, could be reduced. Therefore, there is a need for a system that can determine the type of illumination for an image of a scene.

[0008]

SUMMARY OF THE INVENTION A method is provided for determining a light source type in an image. Using a visible light photosensor with an invisible or ultraviolet (UV) photosensor allows a digital camera to determine the type of light source for a scene.

[0009] Other aspects and advantages of the present invention will become apparent from the following detailed description and the accompanying drawings, which illustrate, by way of example, the principles of the invention.

[0010]

DETAILED DESCRIPTION A system that can determine the illumination type or type of an image of a scene can greatly reduce the time and calculations required to determine the actual light source of the scene. As a result, the white balance of the scene can be quickly obtained using the correct light source.

[0011] Photosensor arrays used in digital cameras are generally red, arranged in a repeating pattern,
Includes green and blue filters. In a typical layout, this repeating pattern has one red, one blue and two
One green photosensor element or pixel.
These four elements make up a superpixel that samples all visible light for this area in the scene of this area. Generally, photo sensors used in digital cameras have low sensitivity to ultraviolet light (UV) (see FIG. 8). By adding an auxiliary UV photosensor to the digital camera, information about the type of light source in the scene can be determined.

FIG. 4 shows a graph of a typical passband filter used for red, green and blue in a typical digital camera photosensor array. Red,
The wavelength range of light sampled using the green and blue filters is typically between 410 nm and 720 nm. FIG. 5 shows an embodiment of the pass band filter of the UV sensor according to the present invention. By using information from the UV photosensor in addition to the red, green and blue elements, the light source type can be determined. In another embodiment (not described in detail here), the UV sensor is 250n
It covers a wider wavelength band from m to 475 nm.

FIG. 6 is a flowchart of a method used in the present invention to determine the type of light source in a digital image of a scene. The operation of each block in the figure is as follows. 602: Measure red light 604: Measure green light 606: Measure blue light 608: Measure ultraviolet light 610: Compare the intensity of ultraviolet light with the total intensity of red, green and blue light 612: Both 614: Ultraviolet light is a fluorescent light source when it is considerably less than the total intensity of red, green, and blue light. 616: Ultraviolet light is compared with the total intensity of red, green, and blue light. 618: UV light is almost the same as the total intensity of red, green and blue light is a sunlight light source 618: In the first step, the intensity of red, green, blue and UV light is measured (602, 604, 606, 608). Then, the intensity of the UV light is compared with the intensity of the red, green, and blue light (610). When the intensity of the UV light is much lower than the intensity of the red, green, and blue light, the light source type is excited to generate light by stimulating a phosphor that emits visible light (which is typically fluorescent). Light source. This is because fluorescent lamps do not produce much light in the UV band. FIG. 7 shows the passbands of the red, green, blue and UV filters superimposed on the power spectrum of a typical fluorescent lamp. The intensity of the fluorescent lamp in the wavelength range of the UV filter is much lower than the intensity over the red, green and blue wavelengths of light. Objects in the scene do not reflect much light in this wavelength range because the light source does not produce much light in the UV band. UV light intensity is red,
If the intensity is lower than the light in the green and blue wavelength ranges, the light source is a tungsten light source (see FIG. 1). If the intensity of light in the UV band is about the same as the intensity of red, green and blue light (or, more generally, both are closer together than with a tungsten light source), the light source type Is one of several spectral curves of sunlight
(See FIG. 2). Once the type or type of light source is determined, the actual light source can be determined more quickly by using a correlation matrix memory method.

Examples of embodiments of the present invention will be listed below.

[Embodiment 1] The following steps (a) to (e)
And determining the type of light source in the scene: (a) measuring light coming from the scene on a first wavelength band; (b) measuring light coming from the scene on a second wavelength band. (C) measuring the light coming from the scene on a third wavelength band; (d) measuring the light coming from the scene on a fourth wavelength band including ultraviolet light; (e) the fourth wavelength. A light source type for the scene is determined by comparing the light intensity of the band with the light intensity of the first, second, and third wavelength bands.

[Second Embodiment] The method according to the first embodiment, wherein the fourth wavelength band passes only ultraviolet rays.

[Embodiment 3] The following steps (a) to (e)
And determining the light source of the scene: (a) measuring the red light coming from the scene; (b) measuring the green light coming from the scene; (c) measuring the blue light coming from the scene; (d) measuring the ultraviolet light coming from the scene; (e) determining the type of light source of the scene by comparing the intensity of the ultraviolet light with the intensity of the red, green and blue light of the scene.

[Embodiment 4] The method according to embodiment 3, wherein the ultraviolet light is measured in a narrow wavelength band.

[Embodiment 5] The following (a) to (c) are provided,
Apparatus for determining a light source type of a scene: (a) a first photosensor array: at least one sensor in the first photosensor array is configured to measure light on a first wavelength band; At least one sensor in the first photosensor array is configured to measure light on a second wavelength band, and at least one sensor in the first photosensor array is on a third wavelength band. (B) a second photosensor configured to measure light on a fourth wavelength band including ultraviolet light; (c) on the fourth wavelength band. The first light intensity;
A processor configured to determine a type of light source for the scene by comparing the intensity of light on the second and third wavelength bands.

[Sixth Embodiment] The first wavelength band is red light, the second wavelength band is green light, and the third wavelength band is blue light. An apparatus according to claim 5.

The foregoing description of the present invention has been presented for purposes of illustration and description. This description is not intended to be exhaustive or to limit the invention to the precise form disclosed, and other modifications and variations are possible in light of the above teaching. By best describing the principles of the invention and its practical applications, those skilled in the art can best utilize the invention in various embodiments and various modifications to suit the particular application intended. Thus, the embodiment has been selected and described. It is intended that the appended claims be construed to include other alternative embodiments of the invention except insofar as limited by the prior art.

[Brief description of the drawings]

FIG. 1 is a graph of a spectral distribution of power of a tungsten lamp.

FIG. 2 is a graph of a spectral distribution of power of D65.

FIG. 3 is a graph of a spectral distribution of power of a fluorescent tube.

FIG. 4 is a graph of a spectral bandpass filter for red, green, and blue filters of a typical photosensor array.

FIG. 5 is a graph of one embodiment of a spectral bandpass filter of a UV filter according to the present invention.

FIG. 6 is a flowchart of a method for determining a light source type of a digital image according to the present invention.

FIG. 7 is a graph of a red, green, blue, and ultraviolet spectral passband filter overlaid on a fluorescent tube power spectral distribution in accordance with the present invention.

FIG. 8 is a graph of the spectral sensitivity of a typical photosensor used in a typical digital camera.

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) // H04N 101: 00 H04N 101: 00 F term (Reference) 2G065 AA06 AB04 AB05 AB18 BA01 BC33 BC35 CA01 DA18 2H002 DB17 2H054 AA01 5C022 AA13 AC42 AC55 AC69 5C065 AA03 BB02 BB48

Claims (1)

[Claims] 1. A method for determining the type of light source of a scene, comprising the following steps (a) to (e): (a) measuring light coming from said scene on a first wavelength band; (b) Measuring the light coming from the scene on a second wavelength band; (c) measuring the light coming from the scene on a third wavelength band; Measuring on the wavelength band; (e) comparing the light intensity of the fourth wavelength band with the light intensity of the first, second and third wavelength bands to determine the type of light source of the scene; Is determined.