EP4169245A1

EP4169245A1 - Methods for producing visual immersion effects for audiovisual content

Info

Publication number: EP4169245A1
Application number: EP21732294.0A
Authority: EP
Inventors: Jérôme DEMOULIN
Original assignee: Cgr Cinemas
Current assignee: Cgr Cinemas
Priority date: 2020-06-18
Filing date: 2021-06-14
Publication date: 2023-04-26
Also published as: MX2022016537A; US20230224442A1; WO2021254957A1; FR3111724B1; FR3111724A1

Abstract

Method for producing visual immersion effects for audiovisual content and sound content associated with the video image, said method comprising the steps of extracting a background of a video image from the audiovisual content; selecting an end zone located at one end of the extracted background; determining a semantic state from the sound content associated with the video image, and processing a predefined image in the selected end zone to generate at least one visual frame intended to be displayed in the peripheral field of vision (9) of a viewer (6) while the video image is being projected in the central field of vision (8) of the viewer (6), the processing of the predefined image being linked to the semantic state determined from the sound content.

Description

Title: Methods for the Production of Visual Immersion Effects for Audiovisual Content

The present invention relates to methods of producing visual immersion effects for audiovisual content such as a film.

[0002] Stimulation of peripheral vision is one of the main factors promoting a feeling of immersion in a viewer placed in front of a screen. It is, in fact, generally admitted that, to have the impression of being in an image rather than in front of an image, the visual field of the spectator must be stimulated in its quasi totality.

[0003] For this, visual frames deduced from the audiovisual content in projection are displayed on either side of a front screen so as to also cover the viewer's peripheral field of vision.

[0004] However, given the specific sensitivities of peripheral vision, particular attention should be paid to the content of these visual frames. Peripheral vision is, in fact, passive and particularly sensitive to contrasts and movements. Unsuitable peripheral content (in strong contrast to what is displayed in the central field of vision, or exhibiting sudden movement for example) can distract the viewer's attention from the video image being projected on the front screen, which reduces, even annihilates, the immersion effect. The content of these visual frames, which escapes the direct analysis of the viewer, must be defined in such a way as to best improve their immersive experience.

As a state of the art, one can take into consideration US2006 / 2683 which describes a method which generates in real time the visual frames intended for the screens based on the content broadcast on the screen and the atmosphere of the room. For this, according to US2006 / 2683, one needs to access the video and audio content for the broadcast.

[0006] According to the invention, the "visual immersion effects" are prepared and constructed upstream of the projection and are not to be confused with the content played on the mediums. These effects form a collection that allows creative teams to generate the final content using the "palette" of effects.

[0007] The generation of this palette allows an enormous saving of time in the production of the contents,

[0008] US2006 / 2683 has the following drawbacks that the method according to the present invention does not present, including the need for metadata to identify image elements, such as the background (see § [0046] [0047] [0061] }, or the need to capture the data stream in real time (see § [0044] [0045] [0048]} since it is not accessible in the context of cinematographic exploitation and the processing is done upstream. Also, this document (see § [0049] - [0052], [0058], [0081] is it of limited use to image projection on physical walls, while according to the invention it is possible to feed display or light devices

The system according to the present invention makes it possible to display visual content on different mediums (for example physical panels, or virtual environment in 3D simulation, etc.) by synchronizing temporally with encrypted media players without having access to the content played.

An object of the present invention is to provide visual frames that best promote an immersive experience based on peripheral visual perception.

[0011] In fact, the invention makes it possible to generate the immersive effects upstream of the projection, which cannot be obtained by the methods and equipment implemented so far.

Another object of the present invention is to generate, for a given audiovisual content, an effects library visual immersion allowing the creation of a visual immersion script for this content.

Another object of the present invention is to be able to automatically generate, for a given film, a visual immersion script intended to stimulate the viewer's peripheral vision during the screening of this film.

To this end, it is proposed, first of all, a method for the production of visual immersion effects for audiovisual content integrating a video image and a sound content associated with the video image, this method comprising the following steps:

- extracting a background from the video image;

- selection of a first end zone located at a first end of the extracted background;

- determination of a semantic state of the sound content; - application of a predefined image processing to the first selected end zone to generate at least one visual frame intended to be displayed in the peripheral field of vision of a spectator during the projection of the video image in the central field of view of the viewer, the predefined image processing being linked to the determined semantic state of the sound content.

Various additional characteristics can be provided, alone or in combination:

- The method further comprises a step of determining a sound parameter of the sound content, the predefined image processing being linked to the determined sound parameter of the sound content;

the sound parameter is chosen from a list comprising a sound height, a sound duration, a sound intensity, a sound timbre, and / or a sound directivity;

- the predefined image processing comprises an adjustment of the colorimetric atmosphere of the first selected end zone;

the predefined image processing comprises a restitution of the average colorimetric atmosphere of the first selected end zone; - the predefined image processing comprises a modification of the brightness of at least one color in the first selected end zone;

- the predefined image processing comprises an application of a blur effect;

the method further comprises a step of selecting a second end zone located at a second end of the extracted background, the second end being opposite the first end, a step of applying said treatment predefined image at the second selected end area;

a plurality of different visual frames integrating said at least one visual frame is generated from the first end zone;

the method further comprises a step of extracting a first plane of the video image, a step of detecting a light beam in the extracted first plane, a step of determining a direction of the light beam detected, a step of generating control data to control a light source capable of generating a light beam in a direction associated with the determined direction;

the method further comprises a step of generating a visual immersion script integrating the visual frame;

- the visual immersion script also includes the order data,

- the command data can be interpreted by a script reader in any form, be it software, hardware, firmware or a combination of these forms.

- the method further comprises a step of adding the visual immersion script to the audiovisual content;

- The method further comprises a step of reading the visual immersion script in a virtual environment in 3D simulation.

Secondly, there is proposed a computer program product implemented on a memory medium, capable of being implemented within a computer processing unit and comprising instructions for implementing the method presented above.

Other characteristics and advantages of the invention will emerge more clearly and in a concrete manner on reading the following description of embodiments, which is made with reference to the accompanying drawings in which:

[0018] the figure [Fig. l] schematically illustrates a video image of audiovisual content;

[0019] Figure [Fig.2] schematically illustrates a background of the video image;

[0020] Figure [Fig.3] schematically illustrates objects of interest in a foreground of the video image;

[0021] Figure [Fig.4] schematically illustrates steps of a method for producing visual immersion effects according to various embodiments;

[0022] Figure [Fig.5] schematically illustrates the stimulation of peripheral vision during the projection of audiovisual content according to various embodiments;

[0023] Figure [Fig.6] schematically illustrates modules involved in the production of visual immersion scripts according to various embodiments.

Referring to Figure 1, there is displayed a video image 1 of audiovisual content intended to be displayed on a front screen 2 disposed facing the viewer. This audiovisual content is, for example, a cinematographic work, or a video film intended to be displayed / projected on a display screen arranged on a front wall of a cinema.

The video image 1 comprises a background 3 (or a background) and a foreground 4 The background 3 corresponds to the scene, the setting or the environment in which one or more are located or evolve. prominent obj ects 41. The first plane 4 comprises objects 41 of interest present or evolving in the environment represented by the background 3. A background 3 is, in fact, generally indexed to the presence of at least one object 41 or a subject, said to be of interest, of foreground on which it is expected that the attention of the spectator. In the absence of foreground objects 41, all of the content of the video image 1 can, in one embodiment, be considered to be the background 3.

The decomposition or the segmentation of the content of a video image 1 in a background 3 and a foreground 4 can be obtained by any method known from the state of the art allowing the extraction of the rear -plane 3 and / or foreground 4. These methods include, for example, background (or, equivalently, foreground) subtraction methods, object extraction methods, search methods moving contours (optical flow or Block-Matching for example), or methods based on deep machine learning (known as “deep learning”). In one embodiment, the extraction of the foreground 4 and / or the background 3 of the video image 1 comprises a step of comparing this video image 1 with the previous and / or following video images of the audiovisual content. .

In one embodiment, the extraction of a background 3 and / or a foreground object 41 from the video image 1 is based on the psychology of the form (better known under the name of "the theory of form" or "the theory of Gestalt") applied to the visual perception of the viewer.

By perceiving the video image 1, the viewer isolates a part which becomes a foreground object 41 on which his attention is focused, and a remainder of the video image 1 which becomes a background 3. The rear -plan 3 is relatively undifferentiated by the spectator who seems to him to extend (by a subjective localization effect) under the foreground objects 41 beyond the contours which limit them or parts of them. This distinction results from the application of one or more laws of Gestalt theory such as - the law of proximity according to which the closest elements in a video image 1 are considered to be perceived by the viewer as belonging to the same group of the foreground 4 or of the background 3;

- the law of similarity according to which the elements having the most graphic similarities (shape, color, orientation for example) in a video image 1 are supposed to induce in the viewer an identical meaning, similar functions or a common importance;

- the law of continuity according to which the more the proximity of certain visual elements in the video image 1 is important, the more they are perceived by the spectator in the continuity as if they are part of the same whole of the background 3 or foreground 4;

- the law of common destiny according to which moving elements following the same trajectory are perceived by the spectator as part of the same whole of the foreground 4 or the background 3.

Thus, the video image 1 is broken down into a foreground 4 and a background 3 (or a background 3). More generally, a first plane 4 and a background 3 are associated with each video image 1 of the audiovisual content.

End zones 31 (or region) are selected from the background 3 of the video image 1. In the example of FIG. 2, these end zones 31 are two zones located at the ends side (right and left) of the background 3. In combination or as an alternative, these end zones 31 may comprise a lower end zone and / or an upper end zone of the background 3 .

An end zone 31 is, in one embodiment, a strip extending from an edge of the background 3 towards its center up to a predefined distance. In another embodiment, an end zone 31 of the background 3 has a generally rectangular shape which covers a border region of the background 3 or, generally, a region comprising an edge of the background 3. the background The dimensions and / or the shape of an end zone 31 can be fixed or variable from one video image 1 to another.

In one embodiment, a first end zone 31 and a second end zone 31 located, respectively, at a first end and a second opposite end of the background 3 (left and right and / or lower and higher for example) are selected.

The selected end zones 31 of the same background 3 can be of different shape and / or dimensions. In one embodiment, two opposite end zones 31 of a background 3 have the same shape and / or the same dimensions. In another embodiment, the selection of a plurality of end zones 31 located at the same end of the same background 3 and being of different shape and / or dimensions can be envisaged.

For example, when the size of the background 3 (or, equivalently, of the video image 1) is 2048x1152, a first left end zone 31 of 0 to 360 pixels by 858 lines and a second right end zone 31 from 1688 to 2048 by 858 lines are selected.

Image processing is applied to each selected end zone 31 of the background 3 to generate visual frames intended to be displayed in the viewer's peripheral field of vision when the image is projected. video 1 in the viewer's central field of vision. This image processing includes the application of graphic effects, cutting operations, cropping (or cropping), non-proportional resizing, and / or geometric transformations (or deformations).

By way of non-exhaustive examples, the graphic effects, generally implemented by means of configurable filters, include

- blur effects such as artistic blur (such as "Bokeh", motion blur, or motion blur), depth of field blur (or background blur), motion blur, radial bum, Gaussian blur, or composite blur;

- sharpness effects such as adaptation of color depth, resolution, definition and / or emphasis;

- colorimetric effects allowing to adapt, to the peripheral vision, for example, the hue of the colors, the light / dark tone, the saturation of the colors, the central color, the correspondence between the colors, the temperature of the colors, the texture , color balance, chromatic replication (or average RGB replication), and / or color curves / histograms of the selected area;

- a modification of the luminosity of at least one color by adapting the contrast, thistogram or the curve of the luminosity, the white balance, the shadows, or the light / dark tone.

[0037] In one embodiment, the image processing comprises an adjustment of the colorimetric atmosphere (via the color balance, or a three-way chromatic corrector for example) of the selected end zone 31. For this, colorimetric effects are applied to the selected end zone 31 so as to generate a visual frame having a certain colorimetric atmosphere. Here, by colorimetric atmosphere (or colorimetry) is meant the general hue that we perceive of a visual frame. This colorimetric atmosphere is, for example, predominantly a predefined color.

In one embodiment, the image processing applied to the selected end zone 31 comprises a restitution of the average colorimetric atmosphere (or average RGB color, i.e. the average of each of the components. Red, Green, Blue)

- of this first selected end zone 31, or

- background 3 of video image 1, or

of the background 3 of the video image let of the background of a next and / or previous video image to the video image 1 in the audiovisual content; Where - of the first selected end zone 31 and of a corresponding first end zone 31 selected from the background of a next and / or previous video image to the video image 1 in the audiovisual content .

In one embodiment, the image processing applied to an end zone 31 is correlated or linked to a sound content of the audiovisual content. This image processing is, for example, linked to a semantic state and / or to a sound parameter of a sound content associated with the video image 1. A semantic state and / or a sound parameter are therefore determined for a content sound associated with the video image 1.

A semantic state of sound content is a semantic description (or description of meaning) of a sound segment. Sound content is able to convey a lot of semantic information. This semantic state is, for example, a meaning attributed to sound content or an expression of feelings / emotions such as joy, sadness, anger, fear, encouragement or, more generally, any event of sound interest. This advantageously results in a visual interpretation of the semantic state of the audio space of the audiovisual content.

A semantic state of a sound content is, in one embodiment, determined following a semantic classification, according to predefined taxonomies, based on sound objects (musical extract, laughter, applause, speech, cry for example) of this sound content and / or a textual description of the sound content (a speech transcription for example). In one embodiment, the semantic classification of the sound content is, moreover, based on a semantic classification of visual objects of the video image 1, in particular the recognition of a visual object of the background 3 and / or a foreground object 41. The recognition of an obj and visual of the video image 1 makes it possible, advantageously, to estimate the source of the sound content and / or the sound context of the video image 1 and, therefore, improve the determined semantic state of the sound content.

In one embodiment, an image processing applied to a selected end zone 31 comprises an adjustment of its colorimetric ambience as a function of the determined semantic state of the sound content associated with the video image 1. By For example, this colorimetric atmosphere is predominantly the color pink when the determined semantic state is “romanticism”, or the color white when the determined semantic state is “happiness”.

The sound parameter is, in one embodiment, chosen from the physical parameters of the sound content integrating a sound pitch (a low / high sound or, more generally, a frequency), a sound duration (a short / long sound ), sound intensity (or volume), sound timbre, and / or sound directivity.

By way of illustration, the image processing comprises the application of a graphic effect correlated with the sound intensity and / or the sound duration of a sound segment associated with the video image 1 in projection . This image processing is, for example, a lighting effect or, in general, a modification of the brightness of at least one color in the selected end zone 31. In one embodiment, the applied image processing comprises changing the degree of brightness of at least one color of the selected end area 31 in proportion to the sound intensity. This makes it possible, for example, to translate a sound peak or a short sound of high intensity (that of a detonation, a shot or an explosion for example) by a visual frame of strong luminosity.

In another example, the image processing comprises an adjustment of the colorimetric atmosphere correlated with the sound height and / or the sound timbre of a sound content associated with the video image 1 in projection. This image processing is, for example, a colorimetric ambience setting for a visual representation of a musical sound (a melody, a rhythm, a harmony, or a certain musical instrument for example) or a voice (voice of a woman or a man).

The image processing applied can also take into account the sound directivity of the sound associated with the video image 1, including in particular the orientation, relative to the viewer, of the visual object supposed to be the source of this sound. and / or its remoteness (intensity). This advantageously results in a visualization or visual interpretation of the sound space of the audiovisual content.

By combining and / or varying one or more image processing applied to an end zone 31 (resizing, filters, intensity, direction or, more generally, of one or more parameters of a processing of image), a plurality of visual frames can be generated for the same selected end zone 31 of a background 3. A non-proportional resizing of the height and / or of the width of the end zone 31 makes it possible to stretch them so as to best cover the viewer's peripheral field of vision.

In order not to expose the viewer's peripheral vision to significant stimulation which could force him to turn his head and, therefore, lose the feeling of immersion, a visual frame is, in one embodiment, of low contrast, low resolution and less sharp than the end zone 31 from which this visual frame is generated. Generally speaking, since the generated visual frames are intended for the activation of peripheral vision, the image processing applied to a selected end area 31 of a background 3 includes a reduction in sharpness. below a predefined threshold.

It results from the image processing applied to an end zone 31 that the generated visual frame comprises one or more indices of the environment at the selected end zone 31 (colorimetry, luminance, appearance , the general shape, and / or the appearance of the objects present in this end zone 31), without however describing it in detail. The display of a visual frame deduced from the video image 1 makes it possible to extend or extend, in the viewer's peripheral field of vision at least partially, the background 3 of the video image

I in projection in his central field of vision. The extension into the peripheral visual field of the background 3 - which constitutes a source of reference points for the viewer in the video image 1 - produces an impression of depth in this video image 1. Indeed, by stimulating the vision peripheral, the latter acts as a vector of perspective which promotes the perception of depth and, consequently, the production of a feeling of visual immersion in the viewer.

During the projection of the video image 1 in the central field of vision of the viewer, the visual frame subjects the viewer's peripheral vision to clues from the background of the video image 1, without however diverting the attention of the front screen viewer 2.

It follows, advantageously, that the visual frame makes it possible to extend the space markers displayed in the video image 1 in order to better still focus the viewer's attention on the foreground objects 41 and on him. provide a feeling of immersion in this video image 1

Advantageously, the visual frame does not include indices of the foreground objects 41 which remain displayed only in the central field of vision of the spectator. The background 3 is extended by these ends to also cover the peripheral field of vision, while the foreground objects 41 remain associated with the central vision. Such occupation of the viewer's visual field advantageously makes it possible to include the viewer in the environment of the video image 1 in projection in his central field of vision and to focus his attention on the front screen 2.

This results for the viewer in an immersive decomposition of the video image 1 in which

- the foreground objects 41 (ie the objects 41 of interest) are subject to his central vision and, therefore, to his direct analysis; and - the setting or environment (i.e. background 3) extends beyond its central visual field to also fill its peripheral visual field.

Each visual frame is intended to be displayed in the viewer's peripheral field of vision on the same side as the end zone 31 from which this visual frame is generated. In other words, each visual frame is intended to occupy a region of the viewer's peripheral visual field. This region diverges from the end zone 31 from which the visual frame is generated.

In another embodiment, the visual immersion induced by the activation of the peripheral vision by means of the visual frames is further amplified by means of an ambient light. This ambient light is emitted by at least one light source capable of emitting a light beam in a predetermined direction. In one embodiment, the hue or the color temperature of the emitted light beam is adjustable. This light source is, for example, a spot, or a directional projector.

The ambient light emitted aims to reproduce a beam of light present in the video image 1 in projection ("flashlight" effect). The light beam present in the video image 1 can correspond to an illumination by a directive light source such as a flashlight, or automobile headlights.

For this, the analysis of the foreground obj ects 41 makes it possible to detect the presence of a light beam in the video image 1 in projection. This detection is, in one embodiment, based on deep machine learning (or, in English, “deep learning”). As a variant or in combination, this detection can be based on the shape and / or the luminosity of the foreground objects 41.

The ambient light control is determined by the direction and the hue of the light beam detected in the foreground 4 of the video image 1 in projection. It is thus possible to reproduce the evolution in successive video images 1 of a light beam produced, for example, by the headlights of a motor vehicle in a bend or by a flashlight being manipulated by a character. In one embodiment, the light beam is reproduced in the vertical peripheral field of vision (in particular, above the central field of vision) of the viewer.

The application of the processing described above to the set of video images 1 of the audiovisual content makes it possible to produce a library of visual immersion effects. This library of visual immersion effects comprises, for each video image 1, one or more visual frames, correlated or not with the soundtrack, and possibly control data of a light source.

This visual immersion effects library constitutes a resource for creating a visual immersion script for audiovisual content. This visual immersion script comprises a series of visual frames and control data from a light source consistent with the initial audiovisual content and intended to be displayed in the viewer's peripheral field of vision during the projection of the audiovisual content.

Indeed, by associating with each video image 1 of the audiovisual content one or more visual frames and, optionally, an ambient light, various visual immersion scripts can be created from this library of effects. visual immersion for the same initial audiovisual content. Each of these visual immersion scripts is advantageously generated natively from the initial source, namely the film or more generally the audiovisual content. This makes it possible to maintain a creative coherence between the choices of the effects constituting the visual immersion script and the initial audiovisual content in its visual and sound narration. A visual immersion script can thus be added to the initial audiovisual content without distorting the initial work.

In one embodiment, a visual immersion script is automatically generated from the visual immersion effects library. For this, a software application (or, of generally, a computer program product) is configured to associate with each video image one or more visual frames and, optionally, ambient light control data deduced from this video image 1. In order to maintain an impression consistent along this visual immersion script, the software application is further configured to guarantee a correlation coefficient between two successive visual frames (visual intra-frames) greater than a first predefined threshold value. This software application is, in another embodiment, configured to choose from among the visual frames associated with a video image 1, one or more visual frames each having, with the end zone 31 from which this visual frame is generated , a correlation coefficient greater than a second predefined threshold value.

This software application is, in another embodiment, also configured to generate, from the audiovisual content, the library of visual immersion effects. In one embodiment, this software application is integrated into a business graphics creation software environment.

The software application is, in one embodiment, able to produce in real time (ie on the fly) from an audiovisual content in projection (in particular, a film) a script of visual immersion intended to be displayed in the spectator's peripheral field of vision at the same time as the projection of the audiovisual content in the central field of vision of this spectator.

In one embodiment, a home cinema system (better known under the name of "home cinema") or, more generally, a television system comprises the software application or a device implementing this software application. This home theater system includes at least a first video output and a second video output arranged to provide a visual immersion script. This visual immersion script is produced in real time by the software application from the audiovisual content in projection on a front screen. This visual immersion script is intended to be displayed on at least two side screens on either side of the front screen. The side screens are, in one embodiment, disposed on the side walls of a chamber.

Referring to Figure 4, the production, from a given audiovisual content, of visual immersion effects comprises, as described above, a step of distinction, for each image video 1 or video shot of this audiovisual content, a background 3 (or background) and a foreground 4. This distinction can result from the extraction of the background 3 (step 10) or the foreground 4. At least one end zone 31 located at one end of the extracted background 3 is selected (step 20). Preferably, two end zones 31 located at two opposite ends, in particular lateral ends, of the extracted background 3 are selected.

The application (step 30) of a predefined image processing to a selected end zone 31 makes it possible to generate at least one visual frame intended to be displayed in the peripheral field of vision of a spectator during the projection of the video image 1 in the central field of vision of the spectator. This image processing adapts the graphic content of the end zone 31 to the viewer's peripheral vision (in terms of sharpness, colorimetry, luminosity, contrast, or dimensions, for example). This image processing is, in one embodiment, linked to the sound content (in its semantic and / or physical dimension) associated with the video image 1 of the audiovisual content. The visual frames thus generated are intended to be displayed / projected on screens aimed at the peripheral vision of the spectator.

Furthermore, the extraction of the foreground 4 makes it possible to detect therein a light beam which, as a result of visual immersion, can be reproduced in the viewer's peripheral field of vision. For this, the direction of this light beam relative to a predefined direction is determined. The hue or the color temperature of this light beam are, in one embodiment, also determined. Control data of a predefined light source for emitting, in the viewer's peripheral field of vision, a light beam in the determined direction or in a direction associated with the determined direction is subsequently generated.

By having the visual frames and the control data thus generated, a visual immersion script for the audiovisual content can be produced.

This visual immersion script can be used in a cinema hall 5, as illustrated in FIG. 5. This cinema hall 5 comprises a front screen 2 and a plurality of side screens 7 on the side on the other side of the front screen 2. The front screen 2 has a screen ratio suitable for covering the central visual field 8 of the spectator 6. As for the side screens 7, they are arranged on the side faces of the cinema hall 5 and are intended to fill the field of view. peripheral vision 9 of the spectator 6. Screens on the ceiling and / or on the floor of the cinema hall 5 (not shown in FIG. 5) can be envisaged to cover the vertical peripheral visual field of the spectator. More generally, any screen making it possible to at least partially fill the peripheral field of vision 9 (horizontal and / or vertical) of a viewer placed in front of the front screen 2 can be envisaged. In one embodiment, the side screens 7 are LED panels.

Eine plurality of light sources 71 capable of emitting a light beam are arranged above the side screens 7, and / or above the front screen 2, on the ceiling, and / or at the back of the room 5 cinema (behind the spectators).

The display of the visual frames of the visual immersion script on the side screens 7 allows an extension or an extension, in the peripheral field of vision 9 of the viewer 6, of the background of the video image 1 in projection on the front screen 2. This produces in the viewer 7 the impression that the background of the video image 1 projected on the front screen 2 is extended into the screens. 7 sides which encompasses it in this video image (an enveloping and immersive effect).

The projection (or display) of the audiovisual content on the front screen 2 and, simultaneously, the visual immersion script on the side screens 7 creates an immersion space making it possible to immerse the viewer 6 in the environment of the scene seen in the video image 1 projected onto the front screen 2. The spectator 6 maintains his central vision on the front screen 2, while remaining aware of who is offered to his peripheral vision by the side screens 7 (comprising, in particular, indices of the environment of the video image 1 in projection).

The visual frames include visual information deduced from the sound content and from the images and video shots of the audiovisual content which are brought into the peripheral field of vision 9 of the viewer 6 to activate / excite his peripheral vision, without diverting his attention from it. front screen 2.

In one embodiment, the immersion script comprises for the same video image 1 a plurality of visual frames intended to be displayed on a plurality of side screens 7 arranged on the same side face of the cinema hall 5 . These visual frames take into consideration the place in room 5 of the cinema of the spectator 6 (placed in the first row, in the middle, or at the back of the room, for example). In another embodiment, these visual frames are increasingly blurred as they move away from the front screen 2 to take into account the fact that the passage from the central zone to the peripheral zone of the field of view is a continuum between the net and the blur and not a straightforward transition. As an alternative, a visual frame is segmented into as many side screens 7, this frame being less and less clear as it moves away from the front screen 2. The number, dimensions and / or arrangements of the side screens 7 are chosen so as to bring the limits of the visual frames closer to the limits of the viewer's field of vision 6 and to leave as little room as possible for real space in his field of vision. . Referring to Figure 6, it is displayed modules involved in the production of a visual immersion script 65 for audiovisual content 61. For this, this audiovisual content 61 is, first of all, provided in input from a generator 62 of visual frames implementing the method described above. Preferably, a plurality of different visual frames is generated for each video image of the audiovisual content 61 so as to obtain at output a palette 63 of immersive effects. Based on this palette 63 of immersive effects, a generator 64 of visual immersion scripts is able to produce one or more scripts 65 of visual immersion.

Secondly, when the audiovisual content 61 is being read by a multimedia player 66 (a software player or a projector, for example), the visual immersion script 65 is simultaneously played by a script player 67. immersion via one or more mediums 68 such as LED panels, a virtual environment in 3D simulation, or display screens.

[0078] The immersive effects palette 63 is, in one embodiment, constructed well in advance of the projection of the audiovisual content 61 by the multimedia player 66, the time to be able to generate several different visual immersion scripts 65.

To ensure synchronous playback of the audiovisual content 61 and the visual immersion script 65, synchronization information is exchanged between the media player 66 and the immersion script player 67. In one embodiment, the immersion script player 67 is separate from the media player 66 so that it does not have access to the audiovisual content 61 being played.

Advantageously, the embodiments described above make it possible to go beyond the two-dimensional display of the front screen 2 by extending the environment / background of the video image 1 in projection to the left. outside this frame so as to activate the peripheral vision of the spectator who, as a result, has the impression of being in the image (a feeling of presence) and not facing a projection on a plane not conducive to a immersion. By being the bearer of meanings, the visual interpretation of the sound content makes it possible to further improve the feeling of immersion in the viewer. In addition, imitation in the viewer's peripheral field of vision of a beam of light present in the projection video image further enriches the viewer's immersive experience.

Claims

1. Method for producing visual immersion effects for audiovisual content integrating a video image (1) and sound content associated with the video image (1), this method comprising the following steps:

- extraction (10) of a background (3) from the video image (1);

- selection (20) of a first end zone (31) located at a first end of the extracted background (3);

- determination of a semantic state of the sound content;

- application (30) of a predefined image processing to the first selected end zone (31) to generate at least one visual frame intended to be displayed in the peripheral field of vision (9) of a spectator (6) ) during the projection of the video image (1) in the central field of view (8) of the viewer (6), this method being characterized in that the predefined image processing is linked to the determined semantic state of the sound content.

2. Method according to the preceding claim, characterized in that it further comprises a step of determining a sound parameter of the sound content, the predefined image processing being linked to the determined sound parameter of the sound content.

3. Method according to the preceding claim, characterized in that the sound parameter is chosen from a list comprising a sound height, a sound duration, a sound intensity, a sound timbre, and / or a sound directivity.

4. Method according to any one of the preceding claims, characterized in that the predefined image processing comprises an adjustment of the colorimetric atmosphere of the first selected end zone (31).

5. Method according to the preceding claim, characterized the predefined image processing comprises a restitution of the average colorimetric atmosphere of the first selected end zone (31).

6. Method according to any one of the preceding claims, characterized in that the predefined image processing comprises a modification of the brightness of at least one color in the first selected end zone (31).

7. Method according to any one of the preceding claims, characterized in that the predefined image processing comprises an application of a blur effect.

8. Method according to any one of the preceding claims, characterized in that it further comprises the following steps

- selection of a second end zone (31) located at a second end of the extracted background (3), the second end being opposite to the first end;

- application of said predefined image processing to the second selected end zone (31).

9. Method according to any one of the preceding claims, characterized in that a plurality of different visual frames integrating said at least one visual frame is generated from the first end zone (31).

10. Method according to any one of the preceding claims, characterized in that it further comprises the following steps

- extracting a first plane (4) from the video image (1);

- detection of a light beam in the extracted foreground (4);

- determination of a direction of the detected light beam;

- Generation of control data to control a light source capable of generating a light beam in a direction associated with the determined direction.

11. Method according to any one of the preceding claims, characterized in that it further comprises a step of generating a visual immersion script integrating the visual frame.

12. Method according to claims 10 and 11, characterized in that the visual immersion script further comprises.

13. The method of claim 12, characterized in that the command data can be interpreted by a script reader in any form, be it software, hardware, firmware or a combination of these forms.

14. The method of claim 11 or 12, characterized in that it further comprises a step of adding the visual immersion script to the audiovisual content.

15. Method according to any one of claims 11 to 13, characterized in that it further comprises a step of reading the visual immersion script in a virtual environment in 3D simulation.

16. Computer program product implemented on a memory medium, capable of being implemented within a computer processing unit and comprising instructions for implementing a method for producing immersion effects visual for audiovisual content according to any one of the preceding claims.