Classifications

• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T3/00—Geometric image transformations in the plane of the image
  • G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
  • G06T3/4023—Scaling of whole images or parts thereof, e.g. expanding or contracting based on decimating pixels or lines of pixels; based on inserting pixels or lines of pixels
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T3/00—Geometric image transformations in the plane of the image
  • G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
  • G06T3/4092—Image resolution transcoding, e.g. by using client-server architectures

Definitions

• the present invention relates to a method dedicated to the resizing of digital images and a system for implementing this method.
• resizing refers to the modification of the resolution of a digital image by enlargement or reduction.
• the field of digital image processing and more precisely the field of digital image resizing, lies in the context of the convergence of fixed, mobile, nomadic, domestic or professional digital imaging technologies.
• a fundamental problem in this field is to meet the increasing demands of the transportability and compatibility of contents between different multimedia devices, the conversion of the same digital contents and their adaptation for different types of media.
• As carriers there may be mentioned mobile phones, PDAs (Personal Digital Assistant), computers, digital television equipment or any other portable or fixed device capable of processing and interpreting digital images.
• complexity of an image processing tool means the consumption of the tool in hardware and computer resources, such as, for example, the size of the operational memory for the execution of the method, the size of the backup memory of the results, calculation power consumption and calculation time.
• US Pat. No. 6,317,524 B1 which relates to a method solely capable of reducing the size of an initial source image. It consists of calculating an error for each pixel and apply differentiated treatment based on the value of this error variable. This solution also requires processing consuming a large amount of computing resources, in particular for calculating the average values of the pixels of the recalculated image as a function of the values of the source pixels.
• patent document WO 2005/055139 A1 Another state-of-the-art reference is patent document WO 2005/055139 A1, which deals with aspects related to image processing, such as optimizing the quality of the image at the time of printing. resizing, is limited only to reducing the size of the digital image.
• patent document WO 2004/057532 A1 discloses a method and a system for reducing and increasing the size of the image by means of insertions or duplications of data by analysis and processing of the diagonals of the image.
• the set of solutions presented has a high complexity of digital processing, which contains convolutions, matrix operations and multidimensional.
• the invention aims to overcome the above-mentioned shortcomings and in particular to design and implement a method and a system with features of speed, low complexity for the different types of resizing digital images without sacrificing the quality.
• the purpose of the present invention is therefore to offer a solution to perform very fast calculations that do not require a large capacity for saving intermediate results or computing power, and this for different types of resizing digital images.
• the invention provides a processing approach on all lines and columns. This pass includes a successive and iterative analysis and calculation by introducing two intermediate dynamic variables.
• the subject of the present invention is a method for resizing digital images consisting of a given column / line proportion comprising at least one digital component organized in rows and columns, characterized in that said digital images are resized by successive and iterative treatments of the row and column position of at least a portion of the digital components of an image, involving at least two intermediate dynamic variables which are functions of the number of rows and columns of a digital image initial and / or number of rows and columns of a resized digital image.
• At least one of the intermediate dynamic variables evolves successively and iteratively as a function of the position of the digital components of the initial image and as a function of the position of the digital components of the resized image, the analysis for the resizing is performed in a single pass on the columns and on the lines of the initializer image on the columns and on the lines of the resized image,
• the resized image, reduced or enlarged has a proportion height / width equal to the proportion height / width of the initial image
• the resized image, reduced or enlarged has a proportion height / width different from the proportion height / width of the initial image
• the resized image has a reduced width and an enlarged height relative to the initial image; the resized image has an enlarged width and a reduced height relative to the initial image,
• the method is applied to at least one digital image format
• the method is applied to at least one digital video format
• the row number values of the initial image and / or column number of the initial image, and / or row number of the resized image, and / or column number of the resized image , relating to the current position of a processed digital element of at least one image are incremented by several units.
• the invention also relates to a system for resizing digital images for the implementation of the method.
• This system is composed of digital processing modules connected to each other, comprising at least one module for resizing digital images by successive and iterative digital processing of the components of the initial digital image using at least one module. calculation and at least one setpoint module for parameterizing the resizing.
• the resized image is encoded in an encoding / decoding module and then saved in a storage module;
• the initial image represents a resized image, processed by the resizing module, directly after at least one resizing or retrieved from the storage module;
• the resized image is displayed on a display module directly after resizing or from the storage module;
• the resized image, displayed on a display module and / or saved by a backup module, is transmitted from a transmission module to a transmission network, and
• the system comprises at least one auto-adaptive setpoint module dedicated to setting the resizing of the initial image.
• the present invention contributes to the convergence of applications, the transportability and compatibility of digital images with different equipment. It also facilitates segmentation into categories for different devices, for the creation of new applications and services according to the requirements of different stakeholders, market needs and the end user.
• FIG. 1 an exemplary diagram of the system according to the invention, composed of digital processing modules connected to each other and of a resizing module
• FIG. 2 a first example of an iterative implementation diagram. implementation of the resizing module according to the preceding figure for the reduction of the digital image by keeping the proportion width / height of the initial image,
• FIGS. 2a to 2c examples of deformation of a square as initial image by variation of the setting values or initial variables
• FIG. 3 a second exemplary iterative diagram for implementing the resizing module according to FIG. 1 for enlarging the digital image while keeping the height / width proportion of the initial image
• FIG. 4 another exemplary iterative diagram of implementation of the resizing module according to FIG. 1 for the reduction of the height and the enlargement of the width of the digital image with respect to the height / width proportion of the initial image
• FIG. 5 a final example of an iterative diagram for implementing the resizing module according to FIG. 1 for increasing the height and for reducing the width of the digital image compared with to the width / height ratio of the original image.
• FIG. 1 An example of a system implementing the method of resizing digital images of the present invention is shown in FIG. 1.
• the system consists of a central unit 72, a display module 98, for example a display screen. computer or mobile phone, and a transmission module 92 which is connected to at least one telecommunications network 88.
• An initial image 10 is transmitted to an input module 18 of the central unit 72 via the link 12.
• Information concerning a digital image 10 to resize, resizing initialization parameters 20, containing the type of resizing and of processing to be applied to this image 10 are also transmitted to the input module 18 of the central unit 72 via the link 14.
• At least a portion of the information acquired by the input module 18 is transmitted to a calculation module 30 via the link 28.
• the calculation module 30 is a digital processor.
• this module can be a microprocessor, a DSP (Digital Signal Processing in English) or any other device capable of performing digital processing.
• DSP Digital Signal Processing in English
• the processor 30 carries out the processing in real time, in particular when an image 10 is downloaded from a remote server and is displayed directly on the screen of the display module 98. This processor 30 can also perform the processing. deferred time, for example when a set of images is processed and saved for later viewing.
• the information acquired by the input module 18 via the links 12 and 14, is transmitted to a storage module 34 via the link 46.
• the storage module 34 is, in the illustrated example, a hard disk in which are stored the image processing modules, namely resizing modules 36, encoding / decoding 62, setpoint 58 and backup modules of the initial images 54 and resized 50, respectively via links 38, 40, 42, 44 and 48..
• initial image is intended to mean an image that will be processed by the resizing module 36, and under the term “resized image” an image resulting from at least one processing of the resizing module. 36.
• the initial image 10, stored in the initial image module 54, is transmitted to an operational memory module 74 via links 56 and 68.
• This operational memory is in the example illustrated a memory type RAM (initials of "Random Access Memory” in English, that is to say "RAM”).
• the initial image 10 can be transmitted directly to the operational memory module 74 from the input module 18 via the link 16, and resized by the resizing module 36, without being stored in the storage module. initial images 54.
• the processor 30 exchanges data with any module of the storage module 34 via the link 32. Following the information received by the input module 18, the processor 30 downloads into the operational memory 74, the data of the encoding / decoding module. 62 via the links 64 and 68, the reference module 58 via the links 60 and 68, and the resizing module 36 via the links 66 and 68.
• the resizing of this image is performed in the operational memory 74, using the resizing module 36 and using the processor 30, according to the parameters provided by the initialization module 20 and recovered by the reference module 58, as explained below.
• the resized image 76 is stored in the resized image storage module 50 via the links 68 and 52.
• the image 76 is stored in the initial image storage module 54 via the images. links 68 and 56.
• the resized image 76 is then transmitted to the output module 80 via the link 78, then redirected to a display module 98 via the link 96 to be viewed.
• the image can also be directed to an external transmission module 92 via the link 94 or to a transmission module 84 internal to the central unit 72 via the link 82.
• the resized image 76 is sent to a transmission network 88, respectively via the links 90 and 86.
• the transmission network 88 can be fixed, mobile, satellite or any other type of transmission. another network capable of transmitting digital data. In this way, the resized image 76 can be transmitted to any device capable of interpreting and displaying this resized image.
• this image can be transmitted directly, or after display by the display module 98, to the transmission module 92 and / or 84.
• FIG. 1 An exemplary implementation of the resizing module 36 is shown in FIG.
• the term "proportion width / height" of the image refers to the proportion number of columns M / number of lines N of the initial image compared to the proportion number of columns m / number of lines n of the resized image.
• width of the image means the number of columns M, m respectively of the initial and resized digital images and the term height, the number of lines N, n respectively of the initial and resized digital images.
• pixel also refers to a component of the digital image defined by its coordinates X 1 Y in the initial image and its a numerical value P, corresponding to a gray or color level, and a component of the digital image defined by its x, y coordinates in the resized image and its numerical value F, corresponding to a gray level or color.
• the coordinates X are relative to the numbers of columns M
• the coordinates Y are relative to the numbers of lines N of the initial image.
• the coordinates x are relative to the numbers of columns m
• the coordinates y are relative to the numbers of lines n of the resized image.
• number of columns m and number of lines n of the resized image is used to denote the number of columns m and the number of rows n that are desired for the resized image following the resizing processing of the initial image. .
• the numerical values M, N, X, Y relative to the initial image and the numerical values m, n, x, y relative to the resized image are natural whole numbers.
• variables X, Y, P refer to the initial image: X and Y represent the position of the pixel in the image and P the numerical value of the pixel, corresponding to a gray level or color.
• variables x, y, F refer to the resized image: x and y represent the position of the pixel and F the numerical value of the pixel corresponding to a level of gray or color.
• the variable "e” and the variable “e '” represent intermediate dynamic variables, dynamic in the sense that they intervene in a decision-test type block, and whose value is recalculated at each iteration, without storing the previous values.
• variable "e” is a linear combination with respect to the variables in lines, the values of N, n, Y, y.
• variable "e '” is a linear combination, with respect to the variables in columns, of the values of M, m, X and x.
• the multiplication by 2 in these formulas can be replaced by a bit shift, and can therefore be executed very quickly on many hardware architectures.
• the first example aims to achieve the reduction of a digital image by keeping the proportion width / height of the initial image.
• the initial image consisting of M. N pixels
• the initial image is reduced to a resized digital image of size mn pixels (M, m being the number of columns and N, n the number of lines, respectively of the initial image and resized), where M ⁇ m and N ⁇ n.
• the method consists firstly in successively and iteratively analyzing the lines of the image and in applying to each analyzed line a calculation processing comprising three main stages I to III, corresponding to the processing of the lines, to the processing of the columns of the image and a final test with respect to the set point for the resized image: the steps I, II and III are carried out respectively by the modules 22, 23 and 24, included in the resizing module 36.
• the method is first initialized in an initialization module 20 in which the setpoint C transmitted from the setpoint module 58 via the links 60 and 68.
• C (Im, R, M, N, m, n) defines the initial parameters, namely:
• Im is relative to the initial image and includes information on the name, the extension of this image and its location (hard disk or remote server or other),
• R represents the choice of the reduction operation
• Step I performed by the module 22 consists of processing the variables X, x and involves the intermediate dynamic variable "e". It comprises the following operations: incrementing X by one unit in module 221,
• test of the value "e '" in the module 223 if "e'" is less than 0, return to the module 221 and the value of X is incremented by one unit; if the value "e '" is equal to or greater than 0, passing to the module 224 in which the value of x is incremented by one unit and the intermediate dynamic variable "e'" is decremented by the value 2.M,
• step M 1 the processing is performed by the module 23 and comprises the following operations: incrementation of Y by one unit in the module 231,
• step III performed by the module 24, the operations relate to the verification of the values of x and y in the module 241: if x is different from m and / or y is different from n, return to step I , for an operation of incrementing X by one unit in the module
• This final module may be the rescaled image module 15, initial image 19, display 16, external transmission 17 and / or internal 18 of FIG.
• FIG. 2 illustrates the low complexity of the module 36 to provide images of reduced size, while preserving the width / height ratio of the initial images.
• a deformation of an initial image in the square format K1 is obtained by decrementing respectively the width m and the height n in the module 235 (FIG. 2).
• the image is resized in the form of trapezoids, respectively T1 and T2, these trapezes being deduced from each other by applying a 2.5D rotation (arrow F1).
• the parallelogram deformation P1 of the square K1 is induced by a variation at the beginning of each line of the initialization value of the X or Y coordinates. This variation is carried out at the initialization step 20 modules 22 or 23 ( Figure 2).
• the invention also makes it possible to use the resizing module 36 as represented in FIG. 3 to achieve, with the same advantages, the enlargement of a digital image while keeping the width / height proportion of the initial image, the setpoint C (Im, A, M, N, m, n) transmits the parameters already described, with a value "A" for the magnification.
• the initial image consisting of M. N pixels
• a final digital image of size m.n pixels, where M ⁇ m and N ⁇ n.
• the method consists first of analyzing successively and iteratively the lines of the image and to apply to each analyzed line a calculation-processing comprising the three main stages I to III. These steps I, II and III are carried out by the same modules, included in resizing module 36, respectively by the module 22, the module 23 and the module 24.
• Step I performed by the module 22 consists of processing the variables x, X and involves an intermediate dynamic variable "e”. It then includes the following operations:
• checking the value of x in the module 326 if x is less than m, return to the module 321 in which the value of x is incremented by one unit. If the value of x is equal to or greater than the value of m, move to module 23 and step II below which processes the variables y, Y and involves a dynamic variable "e".
• step II the processing is carried out by the module 23 and comprises the following operations: incrementation of y by one unit in the module 331,
• test of the value "e” in the module 333 a) if "e" is less than 0, change to the module 334 in which the test of the value y is carried out,
• step III carried out by the module 24, the operations relate to the verification of the values of x and y carried in the module 241: if x is different from m and / or y is different from n, return to the step I, for an operation of incrementing X by one unit in the module 321,
• FIG. 3 describes the process carried out by the resizing module 36 (FIG. 1), a method of low complexity which provides images of increased size, while maintaining the height / width proportion of the images. initial images.
• the present invention also applies to the deformation of images.
• the term "deformation" of a digital image refers to the resizing of said initial digital image into a resized digital image whose width / height ratio is different from the proportion-width / height of the original image.
• FIG. 4 illustrates this application dedicated to producing a digital image deformation by modifying the width / height ratio of the initial image with a reduced height and an enlarged width for the resized image.
• the initial image consisting of M. N pixels
• M being the number of columns and N, n the number of lines
• M> m and N ⁇ n M> m and N ⁇ n.
• the method consists in successively and iteratively analyzing the lines of the image and in applying to each analyzed line a calculation processing comprising the three main steps I to III. Steps I, II and III are performed by the same modules included in resizing module 36, respectively by module 22, module 23 and module 24.
• the method is initialized in the initialization module 20 where the setpoint C is transmitted from the reference module 58 via the links 60 and 68.
• the set point C (Im, HRLA, M, N , m, n) defines the initial parameters as before with, additionally, a parameter H R L A which represents the choice of the operation of reducing the number of lines and enlarging the number of columns.
• Step I performed by the module 22 consists of processing the variables X, x and involves an intermediate variable "e”. It then includes the following operations:
• step II the processing is performed by the module 23 and comprises the following operations:
• step III carried out by the module 24, the operations concern the verification of the values of x and y in the module 241: if x is different from m and / or y is different from n, return to the step I, for an operation of incrementing X by one unit in the module 421,
• This third embodiment describes the process performed by the redimensionrison- module 36 ⁇ ( Figure-1) r-method of low complexity which provides images "deformed” with a reduced width and increased height with respect to the width / height ratio of the initial image.
• the resizing module 36 aims to carry out the deformation of a digital image by modifying the width / height ratio of the initial image with an enlarged height and a reduced width for the resized image.
• the initial image consisting of M. N pixels
• a digital image of size mn pixels M, m being the number of columns and N, n the number of lines
• M M ⁇ m and N ⁇ n.
• the method consists in successively and iteratively analyzing the lines of the image and in applying to each line analyzed a treatment comprising the three main steps I to III. These steps I, II and III are carried out by the same modules, included in resizing module 36, respectively by the module 22, the module 23 and the module 24.
• the method is first initialized in an initialization module 21 with the reference C, transmitted from the reference module 58, C (Im 1 H A L R , M 1 N, m, n). which defines the initial parameters as in the previous example.
• HAL R represents here the choice of the operation of enlarging the number of rows and reducing the number of columns.
• Step I performed by the module 22 consists of processing the variables x, X and involves an intermediate dynamic variable "e”. It then includes the following operations:
• checking the value of x in the module 527 if x is less than m, return to the module 521 in which the value of x is incremented by one unit; if the value of x is equal to or greater than the value of m, move to module 531 and step II below which processes variables Y, y and involves an intermediate dynamic variable "e".
• step II the processing is carried out by the module 23 and comprises the following operations: incrementing the value of Y by one unit in the module 531,
• step III carried out by the module 24, the operations carry as previously on the verification of the values of x and y in the module-241:
• step I for an incrementing operation of X of one unit in the module 521
• This fourth exemplary embodiment describes the process carried out by the resizing module 36 (FIG. 1), a method of low complexity providing "deformed" images with increased height and reduced width, compared to the width / width ratio. height of the initial image.
• the method of the present invention is applicable to more than one digital image format (for example JPEG, JPEG2000, GIF, PNG, BMP, WBMP or other) or digital video.
• the invention is also applicable by successive and iterative analysis of the columns or diagonals of the processed image.
• the values of X, Y, x and y can be incremented by several units according to the intermediate dynamic variables "e” and "e", the values of M, N, m and n and the parameterization of the module. setpoint.
• the method and system of the present invention offer different variations of resizing with or without keeping the width / height ratio of the initial image.
• the method according to the present invention is of low complexity compared to other methods known from the prior art.
• the method according to the invention, of low complexity is suitable for implementation on "light" systems, that is to say systems with constraints of processing capabilities, such as applications embedded on mobile phones or equipment with only limited computing resources.

Landscapes

• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Engineering & Computer Science (AREA)
• Theoretical Computer Science (AREA)
• Image Processing (AREA)
• Editing Of Facsimile Originals (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=39691053

Family Applications (1)

Country Status (3)

Family Cites Families (5)

• 2008
  • 2008-01-30 FR FR0850552A patent/FR2926918B1/fr not_active Expired - Fee Related
• 2009
  • 2009-01-30 EP EP09721298A patent/EP2245590A2/de not_active Withdrawn
  • 2009-01-30 WO PCT/FR2009/000110 patent/WO2009115659A2/fr not_active Ceased

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events