Method for determining distances

Abstract

Method for determining absolute value distances between elements of an image (Im) captured by a monocular camera (C) and said camera, characterized in that it implements the following at a distance determining device (MDD): - receiving (E201) the captured image, - from the received image: -- identifying (E202) elements represented on the image, -- determining (E203) relative value distances, between the elements of the image and the camera, -- identifying (E204) at least one standard element (EE) in said image, -- determining (E205) the absolute value dimensions of said at least one standard element, -- determining (E206), on the basis of the absolute value dimensions of said at least one standard element and the relative value distances between the elements of the image and the camera, the absolute value distances between the elements of the image and the camera. Figure for abstract: Figure 2

Description

Title of the invention: Method for determining distances Field of invention

[0001] This invention relates to the field of image analysis and spatial location.

[0002] More specifically, the invention relates to the determination of distances between elements of an image captured by a monocular camera and said camera. Prior art

[0003] The proliferation of autonomous or semi-autonomous communicating mobile objects (drones, autonomous robots, connected cars) leads to the need to use means of location in the space / near environment and in particular means of estimating distances. This distance estimation can in particular be carried out using cameras / video sensors whose images are subsequently analyzed.

[0004] By determining distances, we will refer here to the distance between the different elements present in the three-dimensional space captured / filmed by the camera and the lens of said camera. Subsequently, we will also speak of the distance between an “element” and a “point of view” or even the “element-point of view” distance. Such an element can refer to a three-dimensional element (for example an object, a person, a structure, etc.), to its representation in a two-dimensional image in the form of a pixel or grouping of pixels. The point of view refers to the position of the lens of the terminal having generated the image (video camera, photo camera, etc.), in the three-dimensional space.

[0005] By extrapolation, this determination of distances can also make it possible to determine the distances between elements of the three-dimensional space and a specific element on which the camera is fixed (for example a car or a drone).

[0006] One of the difficulties in determining distances using cameras is that these distances are measured in three-dimensional space, whereas the images captured by the camera have only two dimensions.

[0007] One solution is to use stereoscopic or rangefinder cameras (Laser, Lidar). Thanks to their two lenses, stereoscopic cameras can capture images of a scene from two different viewpoints. Since the distance separating each lens from the other is known, it is then possible, using stereoscopic measurements, to determine the distances separating the camera from the elements of the scene.

[0008] For their part, telemetric cameras, which are directly designed to measure distances, adequately address this need.

[0009] However, these two types of camera (stereoscopic and telemetric) have the disadvantage of being expensive and relatively rare, which complicates their use.

[0010] At the same time, there are image processing methods for estimating the distances between the elements of a scene and a camera equipped with a monocular lens.

[0011] These methods use algorithms (for example MiDaS or any other algorithm known to those skilled in the art) in the form of neural networks trained to predict the relative distances (to within a scale factor) between the viewpoint and each of the pixels of the image, without even necessarily resorting to semantic analysis of the objects in the scene. These algorithms produce a 2D table of the relative proximities corresponding to each pixel.

[0012] At the same time, there are also other algorithms, such as YOLO (You Only Look Once, Unified, Real-Time Object Detection) or Detectron, which are capable of implementing object detection, with the aim of having a semantic understanding of the scene. These algorithms produce a collection of "detections" each having the following information: type of object, confidence in the detection, rectangle encompassing the shape and possibly masking pixels.

[0013] The elements thus identified on the image are analyzed (vanishing line, point of view / observation, focal length, etc.) in order to determine the distances between the elements of the scene and the lens capturing the scene.

[0014] Although these solutions allow the use of non-stereoscopic cameras, which are more widely used and less expensive than stereoscopic or rangefinder cameras, they also have the disadvantage of only calculating distances in relative value (to within a scale factor) and not in absolute value.

[0015] Here, the expression "distances in relative value" means that the distances between the camera lens and the elements of the captured scene are determined relatively, with respect to each other.

[0016] For example, a relative value distance estimation solution determines that an element A of the captured scene is at a distance x from the objective of the video sensor, while an element B of the scene is at a distance of 2x from this sensor or that an element C of the scene is at a distance of 0.75x from this sensor.

[0017] An absolute value distance refers to a distance established according to a measurement from a unit system such as, for example, the international system (centimeter, meter, etc.), imperial (inch, foot, etc.), etc.

[0018] For example, if x = 2 meters, element A of the captured scene is at a distance of 2m from the lens of the video sensor, while element B of the scene is at a distance 4m from this sensor or that element C of the scene is at a distance of 1.5m from this sensor.

[0019] A preliminary calibration of the “monocular camera + algorithms” solutions makes it possible to determine, by extrapolation, the distances in absolute value between the elements present on the image and the camera lens, on the basis of distances in relative value between elements and lens. However, this process proves to be time-consuming and cannot be deployed in contexts where the cameras or video sensors are in motion.

[0020] Object and summary of the invention

[0021] One of the aims of the invention is to remedy at least one of the drawbacks highlighted by the aforementioned state of the art by proposing a method for determining distances using a monocular camera, in accordance with the invention which will be described below.

[0022] For this purpose, an object of the present invention relates to a method for determining distances in absolute value between elements of an image captured by a monocular camera and said camera, characterized in that it implements the following at the level of a distance determination device: - receive the captured image, - from the received image: — identify elements represented in the image, — determine distances in relative value, between the elements of the image and the camera, — identify at least one standard element in said image, — determining absolute dimensions of the at least one standard element, — determining, on the basis of the absolute dimensions of the at least one standard element and the relative distances between the image elements and the camera, the absolute distances between the image elements and the camera.

[0023] The invention offers the advantage of making it possible to determine distances between elements of a scene using images produced by a standard monocular camera. The invention does not require any particular calibration and can therefore be deployed easily and instantly.

[0024] According to a particular embodiment of the method for determining distances in absolute value, the detection of said standard element comprises a detection of a multitude of identical standard sub-elements constituting said standard element.

[0025] This embodiment advantageously makes it possible to guarantee a more precise determination of the distances in absolute value of the standard element and subsequently of the distances in absolute value of all the elements identified on the image. Many shots take place in environments including standardized elements (ceiling tiles, door frames, floor markings, etc.) whose dimensions are known elsewhere. The use of such standard elements and / or sub-standard elements reinforces the accuracy of distance measurements in absolute value.

[0026] According to a particular embodiment of the invention, the determination of the dimensions in absolute value of said standard element comprises the consultation of a database providing the dimensions in absolute value of the standard element and / or of the standard sub-elements constituting said standard element.

[0027] This embodiment allows the invention to adapt to a large number of environments having different standard elements, by advantageously making use of existing databases listing dimensions of objects / elements in absolute value (databases, parameter files, building plans, etc.).

[0028] According to a particular embodiment, the method for determining distances in absolute value is iterated a plurality of times.

[0029] This embodiment has the advantage of allowing distances to be determined, even when the standard elements move relative to the video sensor, typically when the camera is moving. The invention can then advantageously be implemented in mobile terminals or vehicles equipped with a video sensor.

[0030] The invention also relates to a device for determining distances in absolute value between elements of an image captured by a monocular camera and said camera, characterized in that it implements the following at the level of a distance determination device: - receive the captured image, - from the received image: — determine distances in relative value, between the elements of the image and the camera, — detect at least one standard element in said image, — determining absolute dimensions of the at least one standard element, — determining, on the basis of the absolute dimensions of the at least one standard element and the relative distances between the image elements and the camera, the absolute distances between the image elements and the camera.

[0031] Such a device is particularly suitable for implementing the method for determining distances in absolute value according to any of the embodiments described previously.

[0032] The invention also relates to a computer program comprising program code instructions for implementing the method for determining distances in absolute value as mentioned above.

[0033] Such instructions can be stored permanently in a non-transitory memory medium of a determination device implementing the method for determining distances in absolute value according to the invention.

[0034] This program may use any programming language, and be in the form of source code, object code, or code intermediate between source code and object code, such as in a partially compiled form, or in any other desirable form.

[0035] The invention also relates to a recording medium or information medium readable by a computer, and comprising instructions of a computer program for implementing the method of determining distances in absolute value as mentioned above.

[0036] The recording medium may be any entity or device capable of storing the program. For example, the medium may comprise a storage means, such as a ROM (Read Only Memory), for example a CD ROM (Compact Disc Read-Only Memory), a synthetic DNA (deoxyribonucleic acid) or a microelectronic circuit ROM, or a magnetic recording means, for example a mobile medium, a hard disk or an SSD (Solid State Drive).

[0037] On the other hand, the recording medium may be a transmissible medium such as an electrical or optical signal, which may be conveyed via an electrical or optical cable, by radio or by other means, so that the computer program it contains is remotely executable. The program according to the invention may in particular be downloaded over a network, for example an Internet-type network.

[0038] Alternatively, the recording medium may be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the aforementioned distance determination method.

[0039] According to an exemplary embodiment, the present technique is implemented by means of software and / or hardware components. In this regard, the term "module" or "interface" may correspond in this document to a software component, a hardware component or a set of hardware and software components. Brief description of the drawings

[0040] Other characteristics and advantages will appear on reading particular embodiments of the invention, given as illustrative and non-limiting examples, and the appended drawings, among which:

[0041] [Fig. 1a] represents an example of architecture in which the method of determining distances is implemented.

[0042] [Fig.lb] completes one example of architecture of [Fig.la] in which the method of determining distances is implemented.

[0043] [Fig.2] describes the steps of the method according to one embodiment.

[0044] [Fig.3] illustrates an image from the example architecture presented in figures [Fig.1a] and [Fig.1b].

[0045] [Fig.4] describes the steps of the method according to one embodiment.

[0046] [Fig.5] describes another example of architecture in which the method of distance determination is implemented.

[0047] [Fig.6] describes an example of a distance determination device in one embodiment of the invention. Detailed description

[0048] Description of an example of architecture in which the distance determination method is implemented

[0049] With reference to [Fig.1a] an example of architecture is described in which the method for determining distances is implemented. This architecture comprises: - an environment Env within which we find: — a camera C associated with a PV lens, said camera being placed in a three-dimensional space (shown in section from above), — several elements On i, On, On+[, On+2, constituting the environment Env, - a distance determination device MDD receiving the images captured by the camera C.

[0050] According to the embodiments envisaged, the environment Env is composed of several elements (object, person, wall, floor, ceiling, building, etc.),

[0051] The PV lens is associated with a line of sight LV corresponding to the orientation of the camera + lens pair. The PV lens is also associated with a field of vision (represented by the light zone in the figure) delimited by left field borders BCG and right field borders BCD, as well as by high and low field borders not shown in the figure.

[0052] The elements placed between these field borders from the point of view of the lens constitute elements of a scene captured by the camera C and will therefore be visible within an image captured by the camera.

[0053] The elements On i, On, On+i, On+2 respectively a wall, a table, a chair, a television, are placed at respective distances d(0n4), d(On), d(On+i), d(On+2) from the PV objective.

[0054] According to different embodiments envisaged, the distances d(On_i), d(On), d(On+i), d(On+2) designate the distance between the point of view PV and the closest point of each element.

[0055] According to other embodiments, these distances may designate the distances between the point of view PV and the barycenter of these elements.

[0056] According to different embodiments envisaged, the distance determination device can be physically attached to the camera within a connected terminal (for example a smartphone or a tablet) as in the example illustrated, be integrated into the camera (for example in the case of a connected camera or a video surveillance camera) or even constitute an independent third-party terminal (not shown).

[0057] Description of complementary elements belonging to the architecture in which the distance determination method is implemented

[0058] Figure [Fig.lb] represents other elements belonging to the architecture illustrated in figure [Fig.la].

[0059] This architecture includes, in addition to the elements illustrated in [Fig.1a]: - an On+3 element, corresponding to a ceiling slab, - an On+4 element, corresponding to an assembly of ceiling tiles (including the On+3 element) of 6 tiles by 6, or 36 tiles.

[0060] On+3 and On+4 are respectively at distances d(On+3) and d(On+4) from the objective PV.

[0061] Description of the main actions implemented in the distance determination method

[0062] Figure [Fig.2] represents the steps implemented by the distance determination method, in a particular embodiment of the invention, the method being implemented in an architecture similar to that described in figures [Fig.1a] and [Fig.1b],

[0063] In E201, the distance determination device MDD receives an image Im from the camera C, the image Im being representative of the scene in an environment Env similar to that illustrated in figures [Fig.1a] and [Fig.1b]. According to other conceivable embodiments, the distance determination device MDD can receive video content containing a succession of images Im.

[0064] An example of such an image Im is shown in figure [Fig.3].

[0065] The image Im is generated by the camera C of the figure [Fig. la], which captures elements of the environment Env according to the PV point of view, according to the LV axis, and the framing delimited by the BCG, BCD field borders.

[0066] Depending on the context, this image may constitute a still image or be part of video content.

[0067] Figure [Fig.3] includes the images of the elements On i, On, On+[, On+2, illustrated in [Fig.la] as well as the elements On+3, 0n+4 illustrated in [Fig.lb].

[0068] Referring again to [Fig.2], at E202, the distance determination device MDD identifies elements within the image Im taken by the camera C of the environment Env.

[0069] According to this operating mode, the device identifies the elements On i, On, On+[, On+2On+3 0n+4-

[0070] The number of elements identified is given here as an example. Of course, the number of elements identified may vary depending on the type of environment Env.

[0071] Such a step E202 is carried out by several specific processes (segmentation algorithms, contour detection, neural network, etc.) well known to those skilled in the art.

[0072] According to other embodiments, the identification of the elements relates directly to the pixels (without grouping, segmentation of the image). The identified elements then directly designate the pixels of the image and there are then as many identified elements as there are pixels within the image Im.

[0073] In E203, the distance determination device determines the distances in relative value dVR(On.i), dVR(On), dVR(On+i), dVR(On+2), dVR(On+3), dVR(0n+4) respectively of the elements On i, On, On+i, On+2, On+3, On+ identified on the image in E202.

[0074] Here the term “distance” (relative or absolute) designates the distance between the elements represented on the image and the PV lens of the camera C in the environment Env.

[0075] The determination of distances in relative value can for example be based on an analysis of the vanishing lines of the elements On i, On, On+[, On+2, On+3, On+ identified, making it possible to identify the position from the point of view PV of the objective and its relative distance with respect to the elements On i, On, On+[, On+2, On+3, On+4 identified on the image Im.

[0076] Furthermore, the distance determination device can receive in E201 or in a reception step separate from E201, additional information from the camera C, for example data relating to the focal length of the PV lens or to the inclination of the PV lens relative to the camera C, this data then being able to be used to calculate the distances in relative value between the PV lens / viewpoint and the identified elements.

[0077] According to other embodiments, the distance determination module calculates the relative distance of all the pixels of the image. This step is based on the implementation of algorithms based on neural networks trained to predict the relative distance between the viewpoint and each of the pixels of the image. These algorithms calculate the relative proximity for each “pixel element” of the image Im.

[0078] In E204, the distance determination device MDD identifies / selects a standard element EE within the image Im.

[0079] According to the embodiments, the module selects a standard element from among elements already identified in E202 or identifies the standard element using object detection and / or instance segmentation algorithms, known to those skilled in the art.

[0080] According to certain embodiments, the selection criteria for the standard elements can be fixed / predefined. The distance determination device then identifies one or more specific elements (window frame, ceiling slab, tiling, etc.).

[0081] According to other embodiments, the identification / selection of the standard elements is dynamic. In this case, the distance determination device searches for a standard element EE corresponding to specific criteria (size of the element EE relative to the overall size of the image Im, position with respect to the line of sight, etc.).

[0082] According to possible embodiments, the distance detection device can identify several potential standard elements. In this case, the distance determination module detects / selects from among several potential standard elements that it has preselected, the element that will be designated as being the standard element. The criteria for selecting the standard element can vary according to the embodiments. Among these, the following may be retained in particular: - the recognition of an element within a base of known elements, - image quality criteria at the level of the representative area of ​​the element (overexposure, underexposure, blur, distortion, etc.), - the position of the element relative to the line of sight LV of the PV lens, whether the vanishing lines of the element are visible or not, the angle of the vanishing lines, - etc.

[0083] According to one embodiment, the distance determination device identifies, for example, the element On+3 as being a standard element EE. Subsequently, the element On+3 will be referred to indistinctly as On+3 or EE. Similarly, the distance d(On+3) between the objective / viewpoint and the element On+3 will be referred to indistinctly as d(On+3) or d(EE).

[0084] According to certain embodiments, the distance determining device can identify more than one standard element.

[0085] If the distance determining device fails to identify at least one standard element, the distance determining method ends, and otherwise it continues.

[0086] In E205, the distance determination device MDD establishes a measurement in absolute value MVa(EE) of the dimensions of the standard element EE.

[0087] According to the embodiments envisaged, the determination of the dimensions of the standard element can be based on different approaches including in particular precise recognition of the identified element: - predetermined data (example: one person = lm75, one ceiling tile = 60cm x 60cm, etc.), - manual query / information: according to such an embodiment, a user is invited to enter the dimensions in absolute value of the standard element via a specific interface associated with a third-party terminal or with the camera if the latter has a user interface, - consultation of plans or architectural data referencing for example the dimensions of a wall or the frame of a window of a specific building, etc., - consultation of a database, for example to know the dimensions of ceiling tiles or floor tiles, the dimensions of the grille of a specific car model, etc., - etc.

[0088] In the example shown, the distance determination device establishes that the element EE is a square-shaped slab whose sides measure, in absolute value, 60 cm.

[0089] According to possible embodiments, the distance determination device can receive, prior to step E205, additional data from the camera or a third-party terminal, for example geolocation data, the references of a room or a building, making it possible to facilitate the determination of the dimensions in absolute value MVa(EE) of the standard element EE.

[0090] For example, the distance determination device MDD receives, from the camera C or from a third-party terminal, information indicating that the camera C is placed in a specific room. While it has selected in E204 one of the ceiling tiles as being a standard element, the distance determination device uses this information to search, in a database, the standard dimensions of the ceiling tiles in this room, and thus obtain the dimensions in absolute value of these tiles.

[0091] In one embodiment, the distance determination device MDD can determine an absolute value measurement of the standard element according to a confidence index, this index being able to depend on the relative size of the standard element within the image Im (e.g.: the standard element occupies the entire width of the image Im, the standard element occupies 2% of the surface of the image Im, etc.), the sharpness of the standard element on the image, etc.

[0092] If the distance determination device fails to identify an absolute value measurement MVa(EE) of the dimensions of the standard element or if the confidence index associated with the determination of the measurement of the dimensions in absolute value of the standard element is less than a predetermined threshold, the distance determination method ends, and otherwise it continues.

[0093] In E206, the distance determination device combines the measurement of the dimensions in absolute value MVa(EE) of the standard element EE determined in E205 and the distances in relative value dVR(On i), dVR(On), dVR(On+i), dVR(On+2), dVR(On+3), dVR(On +4) determined in E202 to calculate all the distances (between the viewpoint and the elements) in absolute value.

[0094] According to this embodiment, the distance determination device first calculates the distance in absolute value dVA(On+3) (also designated dVA(EE)) between the standard element On+3 and the objective of the camera / viewpoint PV.

[0095] This first measurement then makes it possible to determine, on the basis of a rule of three / homothethics, the distances in absolute value dVA(Ori J, dVA(On), dVA(On+i), dVA(On+2), dVA (0n+4) between respectively On i, On, On+i, On+2, On+4 and the objective of the camera / point of view PV.

[0096] According to another embodiment, if the distance determination module has calculated the distance, in relative value, of each pixel, the module determines the distance in absolute value for each pixel.

[0097] According to the embodiments, if the distance determination device has determined the dimensions of several standard elements (for example On iet On), the determination of distances in absolute value can be adjusted according to different modalities, for example by averaging the distances in absolute value obtained for each element depending on whether On iet On is used as the standard element or by weighting the distances in absolute value obtained for each element depending on whether On iet On is used as the standard element according to the confidence index associated with On iet On as the standard element, etc.).

[0098] Description of an embodiment of the method for determining distances using a step element composed of step sub-elements

[0099] Figure [Fig.4] represents the steps implemented by the method for determining distances according to another embodiment, the method taking place in an architecture similar to that illustrated in [Fig.1a] and [Fig.1b]. The image Im, captured by the camera C, is moreover identical to the image Im illustrated in [Fig.3].

[0100] Steps E401, E402, E403 are respectively similar to steps E201, E202, E203 of [Fig.2]. For this purpose, they will not be described again.

[0101] In E404 the distance determination device MDD identifies, according to methods similar to those described in E204 and illustrated with reference to [Fig.2], the element 0n+4 as being the standard element EE and the element On+3 as being a standard sub-element sEE. In the example illustrated, the standard element EE is composed of 36 standard sub-elements sEE according to a distribution of 6 standard sub-elements by 6.

[0102] According to other possible embodiments, the standard element corresponds to an assembly of standard sub-elements sEE having other shapes or constituting other elements (for example bricks, tiles). Such sub-elements can be assembled in a variety of different patterns (grid, staggered, etc.).

[0103] In E405, similarly to E205, the distance determining device MDD establishes a measurement in absolute value MVa(EE) of the dimensions of the standard element EE. However, in this embodiment, the distance determining device first determines a measurement of the dimensions MVa(sEE) of the sub-element sEE to then determine, by calculation, the dimensions of the standard element.

[0104] In this embodiment, the sub-standard element sEE corresponds to a square of 60 cm by 60 cm and the standard element EE corresponds to a square of (60 x 6 = 360 cm by 360 cm).

[0105] Step E406 is similar to step E206 described with reference to [Fig.2]. For this purpose, it will not be described again.

[0106] Description of a particular context of use for implementing the method for determining distances

[0107] [Fig.5] illustrates a particular context of use of the MDD distance determination device in the architecture illustrated in figures [Fig.1a], [Fig.1b].

[0108] The architecture of figure [Fig.5] being similar to that of figures [Fig.1a], [Fig.1b], the identical or similar elements which compose it are designated with the same references. Such an architecture is distinguished from that of figures [Fig.1a], [Fig.1b] in that: - camera C is placed on a vehicle V, which moves according to a motion vector (not shown), - the elements On i, On, On+[, On+2, On+3, and 0n+4, respectively designate a first vehicle, a road terminal, a pedestrian, a second vehicle, a ground marking, a set of ground markings.

[0109] From the point of view of camera C, the set of elements On i, On, On+i, On+2, On+3, and 0n+4 each move according to specific vectors (not shown in the figure).

[0110] The embodiment within this architecture corresponds to an iterative implementation of the method for determining distances described with reference to [Fig.2].

[0111] Another embodiment within this architecture corresponds to an iterative implementation of the method for determining distances described with reference to [Fig.4],

[0112] Said methods for determining distances described with reference to figures [Fig.2] and [Fig.4] can be iterated regularly according to a given frequency or be triggered punctually following specific events, for example when the vehicle V exceeds a certain speed or moves in reverse, etc.

[0113] Description of a distance determination device in one embodiment of the invention

[0114] [Fig. 6] shows the simplified structure of a distance determination device MDD corresponding to a particular embodiment of the invention implemented in an architecture as illustrated in [Fig. 1a] and [Fig. 1b] and / or [Fig. 5].

[0115] Such a device comprises, according to the invention, an E / R communication module adapted to receive information from or to the camera C or to a third-party terminal not shown.

[0116] According to a particular embodiment of the invention, the actions executed by the MDD device, within the framework of the implementation of the distance determination method of the present invention, are implemented by instructions of a computer program PG. For this, the MDD device has the conventional architecture of a computer and notably comprises a memory MEM, a processing unit UTR, equipped for example with a processor PROC, and controlled by the computer program PG stored in memory MEM. The computer program PG comprises instructions for implementing the steps of the distance determination method, in particular the aforementioned actions: - receive an image Im captured by the monocular camera C, - from the received image: — identify elements represented in the image, — determine distances in relative value, between the elements of the image and the camera, — identify at least one standard element EE in said image, — determine the dimensions in absolute value of said standard element, — determine on the basis of the dimensions in absolute value of said standard element and the distances in relative value, between the elements of the image and the camera, the distances in absolute value between the elements of the image and the camera.

Claims

Claims

1. Method for determining absolute value distances between elements of an image (Im) captured by a monocular camera (C) and said camera, characterized in that it implements the following at a distance determining device (MDD): - receiving (E201) the captured image, - from the received image: — identifying (E202) elements represented on the image, — determining (E203) relative value distances, between the elements of the image and the camera, — identifying (E204) at least one standard element (EE) in said image, said identification comprising a detection of identical standard sub-elements constituting said standard element, — determining (E205) absolute value dimensions of said at least one standard element, — determining (E206), on the basis of the absolute value dimensions of said at least one standard element and the relative value distances between the elements of the image and the camera,the absolute value distances between the image elements and the camera.,

2. Method for determining distances according to claim 1, in which the determination of the dimensions in absolute value of said standard element (EE) comprises consulting a database providing the dimensions in absolute value of the standard element and / or of the standard sub-elements constituting said standard element.

3. A method of determining distances according to claim 1 or claim 2, wherein the method of determining distances in absolute value is iterated a plurality of times.

4. Device for determining distances in absolute value between elements of an image (Im) captured by a monocular camera (C) and said camera, characterized in that it is configured to implement the following: - receive the captured image, - from the received image: — determine distances in relative value, between the elements of the image and the camera,

5.

6. — identifying at least one standard element (SE) in said image, said identification comprising a detection of identical standard sub-elements constituting said standard element, — determine the absolute value dimensions of said at least one standard element, — determining, on the basis of the absolute value dimensions of said at least one standard element (EE) and said relative value distances between the image elements and the camera, the absolute value distances between the image elements and the camera. A computer program comprising program code instructions for implementing the distance determining method according to any one of claims 1 to 3, when executed on a computer. Computer-readable information medium comprising instructions of a computer program for implementing the method for determining distances according to any one of claims 1 to 3.

Images