EP3216219A1

EP3216219A1 - Method for producing animated images

Info

Publication number: EP3216219A1
Application number: EP14885847.5A
Authority: EP
Inventors: Marc-Eric Gervais Than; Bruno Loubet; Nicolas BESSON; Yves GUIMIOT; Mikael PETITALS; Sebastien ROQUES
Original assignee: Colin Jean-Claude
Priority date: 2014-11-05
Filing date: 2014-11-05
Publication date: 2017-09-13
Also published as: JP2018501693A; WO2016071577A1; US20170337724A1; JP6590925B2; US11195317B2

Abstract

Method for an animation of the multipage type characterized in that each fixed image of the animation has scalability properties; preferably, each of the fixed images is saved as at least two levels, the first level comprising few details and each following level affording further details to the preceding level.

Description

Method for making animated images

The present invention relates primarily to the field of animated multi-page animated images.

There are animation formats of the multi-page animated image type, in particular the one known by the acronym GIF, for "Graphics Interchange Format" (in French, "format d'échange d'images"). "GIF" type animations provide an interesting intermediary between the image and the video: they make it possible to separately store several successive still images, hence the multipage label, and to display them at regular intervals. This makes it possible to have animations requiring computer processing much lighter than a video. These animations are usually played cyclically, so that each image returns to a normally regular interval.

In a video, one can compress the images according to another image, for example by means of predictive functions of the movement, such as in the MPEG family formats. On the contrary, for an animation of the multipage type, each still image is independent of the other still images; thus, the disadvantage of these formats is that, when the size of the images becomes important, the corresponding media is heavy and difficult to transport and shareable. Therefore, an animation consisting of such images, when published on a website or sent to a mobile device, can be slow to load or display. - -

The object of the invention is to propose a method that makes it possible to obtain an animation that incorporates the advantages of animations of the GIF type while being more fluid and quickly downloadable. According to the invention, such a method comprises the use of scalable fixed images, that is to say images having scalability properties; preferably, each of the still images is stored in at least two levels, the first level comprising few details and each subsequent level providing additional details at the level which precedes it Advantageously, to play an image, one reads first a level before the level of higher rank, preferably starting from the first level of the image For each image decompresses and displays the data contained in all the available levels for said image, or, one uncompresses and displays the corresponding data only at the levels available for all the images, or, one decompresses and displays only the data corresponding to the smallest set of levels making it possible to reach at least one given display resolution of the image.

Preferably, for each of the images, each of its levels is located, in the reading order of the media, before the next level. The media of the animation may include, in the playing order of said media, all the levels of a first rank according to the order in which the images of the animation are intended to be played, and then all levels of animation. rank immediately higher, preferably starting with the first levels.

An image may be played according to the data from a level of said image before the data of the higher rank level has been read and processed.

The processed image is advantageously redimaged, before being played, to achieve the desired display resolution.

The still images are advantageously previously compressed by a wavelet method. The image media is preferably a file. In one embodiment of the invention, the still images have been previously compressed by a method comprising at least one wavelet transform.

In one embodiment of the invention, the still images have been previously compressed by a method comprising at least one difference between two adjacent pixels.

In one embodiment of the invention, at least one first still image has been compressed using a first compression method, and at least one second still image has been compressed using a second method. compression.

Advantageously, the method according to the invention comprises a step of analyzing each fixed image in order to determine the most appropriate compression method.

Advantageously, the first compression method comprises at least one wavelet transform, and the second compression method comprises at least one difference between two adjacent pixels. . .

In one embodiment of the invention, the analysis step classifies the images by contrast levels, the least contrasted images being compressed according to the first method, and the most contrasted images being compressed according to the second method.

The invention also relates to a medium for the implementation of a method according to the invention, characterized in that it comprises, in the reading order of said medium, all levels of a first rank according to the order in which the images of the animation are scheduled to be played, then all levels of next higher rank, preferably starting with the first levels.

The invention also relates to a method for selecting a still image of such an animation medium, where a level of a rank is used to display a preview of the media images without reading or without reading and processing the images. higher ranks. .

The invention also relates to a method, characterized in that a fixed image medium is created from the levels of the selected still image.

The invention also relates to a computer program configured, when executed on a processor, to perform a multi-page type animation, characterized in that the animation medium comprises still images having scalability properties, preferably for all the images.

Scalability allows, for the compression formats that use it, to get a preview of the image with only a part of the data and to obtain, by levels, more and more precise previews of the image as and as the data is decompressed. Such scalability is used statically, that is to say for an isolated image, for example in the formats Jpeg2000 and PGF. The GIF format, meanwhile, allows the storage of multiple images within the same media, but not scalability.

A "consistent" set of information created during compression will be referred to as a "level" that, in combination with previous levels, decompresses the image to a defined resolution. Advantageously, this level contains all the layers of the image for this level. For example, if the image is of type YCbCr, the level advantageously contains data for the Y layer, the Cb layer and the Cr layer for this level. Thus, the information contained in a level, combined with information from previous levels, can decompress and display the image at the desired resolution.

Several embodiments of the invention will be described below, by way of non-limiting examples, with reference to the appended drawings in which:

- Figure 1 schematically illustrates the levels of a compressed fixed image, shown in two dimensions. The size of each of the rectangles represents the resolution of the image, once the set of levels up to that represented by the rectangle are decompressed; FIG. 2 illustrates a multi-page animation, comprising five compressed images of three levels of equivalent resolution for each of the images, according to the mode of FIG. 1;

FIG. 3 illustrates the organization of the data of the animation of FIG. 2, according to a first mode of implementation of the invention, where the images are stored one after the other, and by successive levels within each picture ;

FIG. 4 illustrates the organization of the data of the animation of FIG. 2, according to a second mode of implementation of the invention, where the levels of the same rank of all the images are stored one after the other, from the first to the last levels;

FIGS. 5 and 6 illustrate, respectively in a representation in one and two dimensions, a possible display mode of the animation, according to a method of the invention in the embodiment of FIG. 4.

FIG. 1 illustrates the resolution obtained during the decompression of each of three successive levels N1-N3 previously compressed. Thus, the decompression of a single first level NI, comprising the least details, will generate a resolution image such as that of the area occupying the upper left corner of the matrix M. This partially decompressed image can obviously be resized to reach the image display resolution, or the size of a screen for example. The decompression of the data of a second level N2, combined with those of the first level NI, will generate a more detailed image, whose resolution will be equal to the whole area occupying the upper left corner of the matrix M, plus the area intermediate matrix M, located at the bottom right of it. Finally, the decompression of the set of levels NI, N2 and N3, combined with each other, will generate a complete image whose resolution is proportional to the surface of the matrix M.

In the example shown:

the first NI level covers 1/16 of the size of the matrix M, that is to say that if only the first level NI is decompressed, an image is obtained which represents 1/16 of the size of the original image; the data of the first and second levels NI, N2 cover 1/4 of the size of the matrix M, that is to say that if only the first two levels N1, N2 are decompressed, an image is obtained which is 1/4 of the size of the original image;

the data of the first, second and third levels N1-N3 cover the entire size of the matrix M, that is to say that if the three levels N1-N3 are decompressed, the set of the original image;

Each level has data representing the image, with a lower resolution than the next level. Each level uses the data from the previous level at the same time as its own data. Such an image is therefore scalable, it has the property of scalability, that is to say that it is possible to decompress a level after the previous one, without having to decompress the following possible levels.

It is faster to decompress the first level to get a first glimpse of the image rather than waiting for the decompression of the three levels; once the decompression of the second level is carried out, an improved preview of the image can be obtained during the decompression of the third level; this is how the scalability property of such an image is expressed.

In the following, we will describe how we use, according to the invention, the scalability property of the images. FIG. 2 illustrates a multi-page animation A comprising a series of five images P1-P5 compressed according to the mode of FIG. 1, in three levels N1-N3. In the description, for each image Pn of the animation A, the level Nm is denoted Pn: Nm with, in this case, n varying from 1 to 5 and m varying from 1 to 3. The images Pn are a a set of images encoded with scalability, or a set of images that are encoded or re-encoded with scalability for the purposes of the application. They can for example be from a set of JPEG images in a gallery of images or images from a camera.

We will call later "media" all the data to decompress and display an image. This data can be contained in a file, called "file" or "image file". They can also be contained in the memory of a mobile device, for example a telephone or a digital camera, or sent via a network protocol. Figure 3 illustrates the arrangement of an AM media for animation A, according to a first embodiment of the invention. The AM medium comprises five pages AM1-AM5, each corresponding, in the order of their display, to a respective image P1-P5. In each page AMn, the data are classified in the order of the levels N1-N3, so that the data of the first level Pn: N1 of a picture Pn are arranged before those of the second level Pn: N2, themselves before those of the third level Pn: N3. Thus, for each image, each level can be read and decompressed before the next.

For each page AMn corresponds a given display time, corresponding to the rhythm of the animation. This representation by level allows, for each image, to decompress and display only the necessary resolution. For example, if the decompression of the NI and N2 levels of each image is sufficient to display each image at the resolution of the screen of the device on which the animation is played, it is possible, with this representation of the data, to uncompress and display only the NI and N2 levels of each image, saving resources compared to the full decompression of all levels.

In a second embodiment of the invention, the level data is reorganized as illustrated in FIG. 4, in a specific animated AG media, so that all the first NI levels are copied at the beginning. of the medium AG, in the order of the images Pn, followed by the second levels N2 and third levels N3, in the same order. In this arrangement, the set of first levels Pn: N1 are read before the second Pn: N2, then the third Pn: N3.

In this configuration, it is possible to display the first image as soon as the first level P1: N1 corresponding has been read and decompressed, and so on, so that whoever views such an animation, does not need to wait for the complete decompression of the corresponding image. Whenever a picture Pn is intended to be played, it can be played using, if desired, all available data for that picture.

Thus, it is possible to play the animation according to several modes: one can for example decompress for each image to be played all the available levels for this image. This reading mode is represented in FIGS. 5 and 6. In the step illustrated in FIG. 5, the set of first levels Pn: N1 has been read and decompressed, and only the second levels of the first two images P1: N2 and P2: N2 have also been. Then, the animation A can be played as in FIG. 6, the first two images PI, P2 being played in their intermediate resolution, corresponding to the details of the second level N2, the following images P3-P5 are played in their lowest resolution, corresponding to the details of their respective first NI level.

You can also decompress the image using the largest resolution available for all images. For example, if the NI level of all the P images is available, but only a part of the N2 levels, we can only decompress and display the NI level for each image, so as not to cause a resolution change between two images successive. Thus, in the case illustrated in FIG. 5, all the images would be played according to the resolution of the levels N 1. It is also possible to decompress an image until the resolution "useful", that is to say the resolution at least equal to the resolution in which the animation is played. Images of the media can be compressed and decompressed according to different compression methods, as soon as the compression method allows spatial scalability, and where it is possible to combine the information of successive levels to obtain a higher resolution image. important.

In one embodiment of the invention, at least one image Pn can be compressed according to the wavelet method. Wavelet-based compression methods are particularly effective at compressing "natural" images such as photographs. In addition, images compressed using a wavelet-based method generally have several levels that can be organized within a medium according to the invention.

In a second embodiment of the invention, at least one image of the medium is encoded according to a method based on differences between the values of the adjacent pixels. Compression methods based on differences between adjacent pixels are generally particularly effective in compressing high contrast synthetic images. An image compressed using a method based on differences between the values of the adjacent pixels may comprise several successive levels, if the levels represent subsets of increasing size of the pixels of the image, and if the pixels of a level are encoded at least in part from differences with respect to the pixels of the previous levels. We then have several successive levels that can for example be organized according to one of the embodiments described in Figures 3 and 4. In a third embodiment, at least two images are compressed according to two different compression methods. For example, at least one first image may be compressed according to a compression method comprising a wavelet transform, and at least one second image may be compressed according to a compression method comprising a difference between two adjacent pixels.

This embodiment makes it possible to obtain the most efficient compression possible according to the type of image. For example, low-contrast images may be encoded using the wavelet transform method, and the contrast images may be encoded using the method including differences between two adjacent pixels.

In one embodiment of the invention, a method of creating a multipage animation includes a step of detecting the image type in order to apply the most appropriate compression method.

In order to be able to decode each image using the adapted method, the media according to the invention may contain header information indicating, for each image, the compression method having been used.

Time information that can be associated with the media AM, AG, indicating the number of images to be played every second, the animation can be played according to the given time information. Means and steps can be provided for changing the display timing of the media when writing and / or playing the media, in order to be able to use slow motion effects, i.e. idle, or acceleration, or by restoring all or part of each image as soon as the user or an application requires it.

It is also possible, from the data of the animated media AM, AG, to extract the data corresponding to a picture Pn from the images of the animation. Thus, the image Pn can be restored as a still image as needed, in extracting, or reading, and rearranging the Pn: N1, Pn: N2, etc. levels of the image, keeping the compressed data as is.

Rather than being played temporally, in the form of animation, the animated media AG can be used as a directory of images P1-P5. The lower levels NI or intermediate N2 can be used to be displayed as icons, and a user having chosen a picture Pn among the pictures P1-P5, only the image Pn will be completely read and decompressed to be displayed in a fixed manner in its higher resolution, from the information corresponding to the third level Pn: N3.

In one embodiment of the invention, such a medium is used for shooting several continuous images in burst. This capture can be done via a camera / camera, or from successive screenshots. After being captured, it is compressed and stored in several levels. When the capture is complete, the user selects, from the stored images, those that he wishes to transform into an animated image. This selection can relate to all the images, every other image, the second half of the images, etc. The user can associate a playback speed with this medium, which can correspond to the capture speed, but also at a faster or slower speed to create "slow motion" or acceleration effects. The application then creates, from the compressed images, an image media organized by levels, such as the animated media AG of FIG. 4. An image extracted from the media can also be stored independently, in the form of a media containing a single image. , fixed.

Of course, the invention is not limited to the examples which have just been described. Thus, the number of levels is not limited to 3, it can be 2 or 4 or more than 4; thus, one can thanks to the method according to the invention, display quickly - - images of a size progressively much larger than those generally used in the GIF, in particular by increasing the number of levels. The invention therefore proposes methods for compressing and decompressing moving images using scalability. These methods allow the animation receiver to display it from data that has already arrived, without necessarily having received everything.

Claims

A method for performing an animation (A) of the multipage type, characterized in that the medium (AM, AG) of the animation comprises fixed images (Pn) having scalability properties.

2. Method according to claim 1, characterized in that each of the scalable images is stored in at least two levels (Nm), the first level (NI) comprising the least details and each subsequent level (Nm) providing additional details. at the level (Nm-i) that precedes it.

3. Method according to claim 2, characterized in that to play an image (Pn), one reads first a level (Nm) before the level of higher rank (Nm + i), preferably starting with the first level of the image (Pn: Nl).

4. Method according to claim 3, characterized in that for each image (Pn) decompresses and displays the data contained in all available levels for said image (Pn).

5. Method according to claim 3, characterized in that one uncompresses and displays the data corresponding only to the levels available for all the images.

6. Method according to claim 3, characterized in that it decompresses and displays only the data corresponding to the smallest set of levels to achieve at least a given display resolution of the image.

7. Method according to one of claims 1 to 6, characterized in that, for each of the images (Pn), each of its levels (Pn: Nm) is located, in the reading order of the media, before the level next (Pn: Nm + i).

8. The method as claimed in claim 7, characterized in that the medium (AG) of the animation (A) comprises, in the reading order of said medium, all the levels (Nm) of a first rank (m) according to the order in which the images (Pn) of the animation are intended to be played, then all levels of immediately higher rank (Nm + i), preferably starting with the first levels (Pn: Nl).

9. Method according to one of claims 1 to 8, characterized in that an image (Pn) is played according to the data from a level of said image (Pn: Nm) before the data of the higher rank level (Pn: Nm + i) were read and then processed.

10. Method according to one of the preceding claims, characterized in that the processed image is resized, before being played, to achieve the desired display resolution.

I L Method according to one of claims 1 to 10, characterized in that the still images have been previously compressed by a method comprising at least one wavelet transform.

12. Method according to one of claims 1 to 10, characterized in that the still images have been previously compressed by a method comprising at least one difference between two adjacent pixels.

13. Method according to one of claims 1 to 10, characterized in that at least a first fixed image has been compressed using a first compression method, and at least a second fixed image has been compressed to using a second compression method. 14. The method of claim 13, characterized in that it comprises a step of analyzing each fixed image to determine the most appropriate compression method.

The method of claim 13, characterized in that the first compression method comprises at least one wavelet transform, and the second compression method comprises at least one difference between two adjacent pixels. 16-A method according to claims 14 and 15, characterized in that the analysis step classifies the images by contrast levels, the least contrasted images being compressed according to the first method, and the most contrasted images being compressed according to the first method, the second method.

17. Method according to one of the preceding claims, characterized in that the image medium is a file.

18. Media (AG) for the implementation of a method according to one of the preceding claims, characterized in that it comprises, in the reading order of said media, all the levels (Nm) of a first rank (m) according to the order in which the images (Pn) of the animation are intended to be played, then all rank levels immediately higher (Nm + 1), preferably starting with the first levels (Pn : Nl).

19. A method for selecting a still image of an animation medium (AM, AG) according to claim 18, characterized in that a level of one rank (m) is used to display a preview of the media images. (AG) without reading or reading and processing images of higher ranks (> m).

20. The method of claim 19, characterized in that a fixed image medium is created from the levels of the selected still image.

21. A computer program configured for, when executed on a processor, performing an animation (A) of the multipage type, characterized in that the media (AM, AG) of the animation comprises images (Pn) fixed having scalability properties, preferably for all of the images.