FR3054707A3 - METHOD FOR ACQUIRING COLOR IMAGES UNDER INCOMING AMBIENT LIGHTING - Google Patents

Abstract

We compare two successive acquisitions of a multicolored object that may be a barcode, one of which is made under ambient lighting that may be unknown, and the other by adding a substantially white light whose spectrum is known. By comparing these two acquisitions, we can obtain the effect of this known lighting, which allows to deduce the colors of the modules constituting the multicolored barcode and thus to decode.

Description

Method for acquiring color images under unknown ambient lighting Area of application The invention is a method for acquiring color images, which is particularly applicable to the reproduction of color photographs and to the use of images. Optical reading codes.

An optical read code is a symbology, that is, a graphical representation of a set of data. Various formats have been proposed for coding how to produce an image that corresponds one-to-one with input data. These formats include: Datamatrix (ISO / IEC 16022), Maxicode (ISO / IEC 16023), QR-Code (ISO / IEC 18004), among others.

Optical codes are composed of a set of lines or points called elementary cells considered as dark or clear according to a rule chosen by their designer. Problem

When photographing a colored object, the colors observed depend on the spectrum of the light source illuminating the object.

When one is in the dark, it is usual to use a white light so that each color of the object is reflected. On the other hand, if one uses a light which does not have the same level of each of the radiations composing the white light but an irregular spectrum, the reflected colors are not those which would have been reflected by a white light. For example, if we use a light source whose spectrum essentially comprises radiations corresponding to the colors red, green and blue, that is to say a higher order white, we have the impression of having a white light but the colors of the object that are not part of these three sets of radiation are not reflected. This results in an error in the appreciation of the colors of the object.

This considerably limits the applications of color barcodes, which make it possible to increase the number of possible combinations of a barcode, without increasing the surface area they occupy, since the spectrum of the barcode is generally not controlled. ambient lighting, or we do not know if we do not have a spectrometer.

Prior art

Inaccurate color acquisition is often unimportant, but there are many cases where it is desired that the measured color be as accurate as possible. This is the case when reproducing the result of a spectrographic analysis, for example, but the most common case is the realization of color barcodes, which have the advantage of allowing the representation of more information on a specific object. given surface. There have been numerous attempts to measure the color of the elementary cells of a barcode, but none have succeeded because the ambient lighting is not always white, which can distort color appreciation.

The original barcode patent - US 2612994 of October 20, 1949 - was filed by two American students, Norman Joseph Woodland and Bernard Silver, and is known from the patents of Wei Ming (Konika Minolta) US7823797B2 and US8215556B2, which describe color barcodes with a color palette used

The following three patents are also known: US 2002/066786 Al (Saito Takahiro [JP] Jun 6, 2002 (2002-06-06), US 2013/126618 Al (Gaop Wenliang [US] May 23, 2013 (2013-05-23 ) and WP01 / 24106 Al (Sicpa Holging SA [CH]) April 5, 2001 (2001-04-05), all of which propose to illuminate a multicolored bar code successively by several monochrome light sources, in order to show each time the one of the corresponding colors of the bar code considered.

None of these solutions makes it possible to acquire a color image when only ambient lighting is available whose photometric spectrum is not known, and a substantially white light produced by a flash such as that of a camera. DESCRIPTION OF THE INVENTION The invention is a method of acquiring an image of a multicolored object, characterized in that it consists in: - acquiring successively with an acquisition means such as a camera two views of the multicolored object, by varying the lighting, one of the acquisitions being made under ambient lighting and the other by adding to this ambient lighting a so-called additional multi-chromatic light source, whose spectrum is known, which covers substantially the entire spectrum of the visible spectrum, such as a camera flash, and comparing the two acquisitions to deduce the effect of adding said additional light source.

According to other characteristics, the value retained for each pixel of an image obtained during an acquisition is a weighted average of the brightness of this pixel and of adjacent pixels; the value retained for each pixel of an image obtained during an acquisition is increased or decreased by an identical coefficient for all the pixels of this image, such as a surface of the multicolored object whose luminosity and / or color are known to be represented at the brightness and / or the corresponding known color; the values retained for each pixel of one of the two images obtained is increased or decreased by an identical coefficient for all the pixels of this image, this coefficient being such that no pixel of the image made with the source of Additional lighting is noted as brighter in the acquisition made with the additional light source than in the other picture with the only ambient lighting. the image obtained by said comparison of the two acquisitions is corrected to take account of the known nature of the photometric spectrum of said additional light source; said multicolored object is a barcode; the method comprises a step which is the monochromatic analysis of the multicolored bar code, leading to a decoding of information, and another step which is the analysis of the color of each of the elementary cells of the multicolored bar code leading to another series of information. said other step also comprises the analysis of the brightness of each of the elementary cells of said multicolored bar code.

Brief description of the drawings The invention will be better understood, and other objects, advantages and characteristics thereof will appear more clearly on reading the description which follows, which is illustrated by the appended figures numbered from 1 to 11.

Figures 1 and 2 show two different bar codes.

Figure 3 shows a combination of these two barcodes, comprising elementary cells of four different colors. These colors are represented by different gray levels.

Fig. 4 shows the secondary cells each containing color samples used for the bar code of Fig. 3.

Fig. 5 shows the combination of the primary elementary cells of Fig. 3 and the secondary elementary cells of Fig. 4.

Figures 6 and 7 show a series of 9 elementary cells of a code, whose primary elementary cells have been colored (the colors are here also represented by gray levels). Figure 7 shows an example of distribution of secondary cells how to have additional information.

Figures 8-11 illustrate one way of adding an image to a code, using the secondary elementary cells. Figure 8 shows in grayscale a color image that will be combined with the enriched code. FIG. 9 represents the parts of this image that will constitute the secondary elementary cells of the enriched code. Figure 10 shows the enriched code. Figure 11 shows the software extracted image of the enriched code.

FIG. 12 shows the result of the printing on a white substrate, by subtractive synthesis, of three surfaces with inks of cyan, magenta and yellow colors, noted respectively c, m and j, which gives rise to the appearance of 7 colors, namely red, green, blue and black, denoted respectively r, v, b and n in addition to the colors of the inks.

FIG. 13 shows the result of the printing on a white substrate, by subtractive synthesis, of four surfaces with inks of cyan, magenta, yellow and black colors, respectively noted c, m, j, and n, which gives rise to 14-color, dark cyan, dark magenta, dark yellow, dark black, dark red, light red, dark green, light green, dark blue, light blue, noted respectively C, M, J, N, R , r, V, v, B and b, in addition to the colors of the inks.

Detailed description of the invention

The principle of the present invention is to compare two successive acquisitions of an object such as a barcode, one of which is made under any lighting, the other being made after adding a substantially white multi-chromatic lighting whose spectrum is known. By comparison, one can thus obtain the effect of this substantially white illumination whose spectrum is known, which allows to deduce the colors of the elementary cells constituting the multicolored barcode and thus to decode it.

There are four main advantages to using on the one hand ambient light and on the other hand the sum of this ambient light and a substantially white multi-chromatic lighting: 1. this allows to use barcodes containing any what color is visible to the human eye 2. it allows to operate without knowing the spectrometry of the ambient lighting; 3. the simple flash of a camera can be used as additional light; 4. and this makes it possible to know exactly the colors in the case where the spectrometry of the substantially white multi-chromatic lighting is known.

Corrections must be made to the results of each of the two acquisitions before subtracting, to account for some of the noise that can disrupt the acquisition and for the change in aperture and time exposure between the two acquisitions, especially when the acquisition means is automated as is very often the case when this means of acquisition is a smartphone.

To reduce the noise, the simplest solution is to correct the value of each pixel of the image obtained by using instead of the original value a weighted average of the brightness of this pixel and neighboring pixels. This operation can be done successively or simultaneously in the three elementary colors of the sensor. It is also possible to reduce the resolution of the image, which produces substantially the same effect. Those skilled in the art can correct the change in the aperture of the lens and / or the exposure time between the two acquisitions when it knows what the parameters have been changed, but sometimes it is not known not what are the changes that have been made by the acquisition device when it is automated. This is the case for many smartphones. In this case, it is advantageous to use one or both of the following two methods: 1. to correct the brightness of the pixels of an image so that a surface whose brightness is known is represented at the corresponding brightness - this operation is advantageously performed on two colors that can be black and white; 2. correct the brightness of the pixels of one of the two images obtained, so that no pixel of the image made with the additional lighting source is noted as brighter in the acquisition with the additional light source than in the other image with the only ambient lighting.

We can use the image obtained by comparing the two acquisitions as it is, if we consider that the difference between the light produced by the additional light source has a sufficiently regular spectrum to produce a truly white light, but the One skilled in the art can easily, knowing the spectrum of this additional light source, correct the image obtained by the comparison to obtain a faithful representation of the colors of the multicolored object.

When the multicolored object is a barcode, the acquisition is advantageously in two steps. We begin with a monochromatic analysis of the multicolor bar code, which leads to a first information decoding. Then, a second step is to analyze the color of each of the elementary cells of said multicolor bar code to obtain a second set of information. Said second step may also include analyzing the brightness of each of the elementary cells of said multicolor bar code. The order of these two steps can be reversed.

It is advantageous to vary the shape and / or the color, and / or the texture and / or brightness of at least one of the dark or light elemental surfaces of a bar code, these variations in shape, and / or color, and / or texture and / or brightness representing additional data. It is particularly advantageous that the light elemental surfaces remain clear and the dark elementary surfaces remain dark, since these elementary surfaces remain considered light or dark by the conventional bar code reading algorithm. In this case, the variation in shape, and / or colors, and / or texture and / or brightness does not interfere with the acquisition of the bar code that can be made during a first step. This first step is followed by a second step in which the shape, the color and the brightness of the primary elementary surfaces are analyzed more finely.

A bar code obtained by this method may additionally comprise one or more elementary surfaces called secondary elementary surfaces containing additional data represented by the shape, and / or by the color, and / or by the texture and / or the brightness of the said elements. secondary elementary surfaces. These secondary elementary surfaces can be analyzed after the first step, to provide a large number of additional data. Those skilled in the art can for example divide the primary elementary cells into several parts, and the secondary elementary cells also into several parts, the whole of each of the same parts of the elementary cells being able to be acquired by a camera and constitute a additional barcode. By dividing a primary elementary cell into four parts and a secondary elementary cell into 5 parts, which corresponds to dividing an elementary surface into 9 parts, there are thus 9 barcodes which each have the same number of elementary cells.

It should be noted that the bar code represented by primary elementary cells may constitute one of these bar codes, and that it remains very easy to analyze by the most conventional algorithm during the first step. The analysis of the 8 or 9 other barcodes requires an analysis done at a resolution three times higher, but it does not need to be done as quickly as that of the first barcode.

Advantageously, a so-called primary primary surface is the central part of one of the elementary surfaces of a set of elementary surfaces constituting a barcode. The person skilled in the art knows how to analyze a barcode by starting by looking at the center of each elementary cell and ignoring the rest of each cell. A secondary elementary surface may therefore consist of at least part of the peripheral part of the elementary surface considered.

As illustrated in FIG. 7, a so-called secondary elementary surface may consist of all or part of the peripheral part of the elementary surface considered, and of at least part of the peripheral part of an adjacent elementary surface. This makes it possible to have larger elementary surfaces, thus easier to acquire by a camera. To avoid that irregular illumination of the code disrupts the acquisition of a code enriched according to the invention, it is recommended to have the color samples, and / or texture and / or brightness near the code.

In a particular implementation illustrated in Figures 4 and 5, these color samples, and / or texture and / or brightness are arranged on the secondary elementary surfaces of the enriched code. This has the advantage of being able to compare the color and / or texture and / or brightness of a primary elementary cell with a sample located very closely and therefore most likely subjected to the same lighting.

It is particularly advantageous in some applications that part of the primary and / or secondary elementary surfaces represents encrypted data. The key to decrypt them can be contained in information represented by other primary and / or secondary elementary surfaces. It can also be present on a server at an address contained in other elementary surfaces. Those skilled in the art can provide all kinds of strategies to communicate the key only to people able to practice decoding.

In a particular embodiment illustrated in FIGS. 8 to 11, the color and / or the brightness of a secondary elementary surface is a representation of the color and / or brightness of a pixel of an image which can therefore be be extracted from said enriched code. The secondary elementary surfaces can be obtained very simply by masking the elementary surfaces of the image, as illustrated in FIG. 9, but those skilled in the art can also calculate the brightness and the color of these surfaces so that they represent at best the brightness and color of the preserved parts and the masked parts.

To read such a code, one operates according to one of the methods described above, and then the photograph is extracted from the set of secondary elementary surfaces. Missing areas (those corresponding to the primary elementary cells) are replaced by the calculation based on the brightness and color of the nearest cells. An attenuation by blurring allows in fine to reconstitute the original image as shown in Figure 11. The original image may have been encrypted before being combined with a barcode, and decrypted during the acquisition of the image. code. It should be noted that the present invention thus makes it possible to display an image as a photograph only to persons having the key, and to do it in local mode, without any connection with an external network.

Advantageously, said acquisition means is placed in such a way that said additional light source is not reflected directly by the printing surface of said barcode towards said acquisition means. Otherwise, this reflection may interfere with the acquisition by saturating the image of the barcode with light.

Main advantages of the invention over the prior art

It allows to acquire images of multicolored objects even under lighting conditions are not known.

In particular, it allows the design of multicolored barcodes that can be read in virtually any lighting condition. Such bar codes allow a number of possible combinations equal to CN for a code consisting of C different color codes having N elementary cells, instead of 2N for traditional monochrome codes. Applications The invention applies to all photographic acquisitions, and in particular to all applications of barcodes. The serialization of the products, which requires the printing of very large numbers in the form of barcodes, is one of the great applications of the present invention.

Claims (8)

claims 1. A method of acquiring an image of a multicolored object, characterized in that it consists in: acquiring successively with an acquisition means such as a camera two views of the multicolored object, by varying the lighting, one of the acquisitions being made under ambient lighting and the other by adding to this ambient lighting a so-called additional multi-chromatic light source, whose spectrum is known, which covers substantially the entire spectrum of the visible spectrum, for example a camera flash, and comparing the two acquisitions to deduce the effect of adding said additional light source, 2. Acquisition method according to claim 1 characterized in that the value used for each pixel of an image obtained during an acquisition is a weighted average of the brightness of this pixel and adjacent pixels. 3. Acquisition method according to any one of claims 1 and 2 characterized in that the value used for each pixel of an image obtained during an acquisition is increased or decreased by an identical coefficient for all the pixels. of this image such as a surface of the multicolored object whose brightness and / or color are known is represented at the brightness and / or the corresponding known color. 4. Acquisition method according to any one of claims 1 to 3, characterized in that the values retained for each pixel of one of the two images obtained, is increased or decreased by an identical coefficient for all pixels of this image, this coefficient being such that no pixel of the image made with the additional lighting source is noted as brighter in the acquisition made with the additional light source than in the other image with the only illumination ambient. 5. Acquisition method according to any one of claims 1 to 4 characterized in that the image obtained by said comparison of the two acquisitions is corrected to account for the known nature of the photometric spectrum of said additional light source. 6. Acquisition method according to any one of claims 1 or 5 characterized in that said multicolored object is a multicolored barcode. 7. Acquisition method according to claim 6 characterized in that it comprises a step which is the monochromatic analysis of said multicolored bar code, leading to a decoding of information, and another step which is the analysis of the color of each of the elementary cells of said multicolor bar code leading to another set of information. 8. Acquisition method according to claim 7 characterized in that said other step also comprises the analysis of the brightness of each of the elementary cells of said multicolored bar code.