FR3075371A1 - SYSTEM FOR AIDING THE DETECTION OF A SURFACE FAULT - Google Patents

Abstract

 The invention relates to a system for assisting in the detection of a defect (DF) in the surface of a wall (PLV) of an object (V). Such a system comprises one or more capture devices (1, 1 '), comprising a matrix sensor (C) and a first, or even a second, light element (EL1, EL2) for projecting onto the wall (POS) of said object ( V) respectively a first heterogeneous motif and a second homogeneous motif. Such capture devices (1, 1 ') cooperate with a processing unit (11) arranged to produce from digital representations (R (t), RW (t)) delivered by said matrix sensor (C) of each device capture (1, 1 ') a composite or fused representation of said wall (PLV) highlighting a possible surface defect (DF).

Description

The invention applies more particularly to the total or partial audit of a bodywork or a chassis, of a land vehicle, such as an automobile, arranged to transport people or goods, or even a accessory of such a vehicle such as a trailer for example.

The term “bodywork” of a land vehicle, in particular and in a non-exhaustive manner, a front door, or even a rear door if there is one, a bonnet, a luggage or service boot cover, fenders, bumpers, etc. The invention can be applied more generally to the capture of views, sections or walls of a mobile object relative to the system for assisting in the detection of a surface defect. As such and within the meaning of the invention, the concept of relative movement between an audited object and a system for assisting in the detection of a surface defect in accordance with the invention translates a displacement of said object with regard to said system or a displacement of the assistance system for the detection of a surface defect, or even a subset of said system, with regard to said object.

The invention applies particularly, preferably, to an audit aimed at detecting a possible defect in the surface of an external envelope of an object, when the latter is a rental vehicle when the said vehicle is made available and / or when the latter is returned. A second application may consist of an easier measurement of damage sustained by a vehicle during its use or during the

following a disaster, such as a traffic accident for example.

To carry out such an audit, at present, one to several professionals are required to visually analyze the body or chassis of a vehicle. Depending on the case, lifting equipment may be necessary to carry out an examination of a vehicle chassis. Detecting one or more surface defects, that is to say an alteration of the paint or varnish covering the bodywork, or even a deformation of said bodywork, may prove to be time-consuming or random depending on the extent of said defect or the expertise Staff. Carrying out a complete diagnosis of a vehicle generally requires immobilizing the vehicle or conveying it to an expert site. Such provisions are generally contradictory with a smooth and rapid return of rental vehicles. For the staff of a rental agency, a rapid visual examination may sometimes prove to be insufficient to detect body defects that are difficult to detect with the naked eye depending on the conditions under which the examination takes place, such as the quality of the ambient lighting.

In addition, disputes may arise between landlords and tenants, debating a pre-existing defect or arising from being frequent between the occurrence of a rental. Indeed, a contradictory examination of a vehicle objectively observing the state of the bodywork prior to rental is impossible to conduct in a few seconds. A tenant can thus commit, when he perceives a vehicle, to a condition of the bodywork not reflecting all of the previous damage suffered during previous rentals, and conversely, a rental company can accept a vehicle without easily and objectively noting new damage caused by the tenant returning the said vehicle.

The invention makes it possible to offer a valuable aid to the visual detection of damage, highlighting any surface defect, whether it is a plastic deformation or a change in color, thus solving all or part of the drawbacks of known solutions. It also allows the implementation of automatic methods of detecting such additional defects aiming on the one hand to detect such a defect or on the other hand to compare different states of a surface over time.

Among the many advantages provided by the invention, we can mention that:

a system for assisting in the detection of a surface defect in accordance with the invention may be of modular structure so as to be able to simultaneously analyze one or more walls of an object in motion relative to said system, during a single passage of the object before said system or vice versa;

- the light and economical structure of such a system makes it possible to equip any rental center for objects, whether terrestrial, when said object is altered when it is returned;

- said system can visually detect, in particular vehicles, the able outer envelope and requests an available audit and easily facilitate or even automatically, surface defects, that is to say a volume deformation following a damage and / or chromatic alteration of said surface as a result of scratching, scratching or even of a surface removal or deposition of a paint or varnish;

- such a system makes it possible to associate one or more digital representations of surfaces of an object with an identification of said object or any other representation of the results of a functional and / or structural control complementary to said object.

To this end, there is first provided a device for capturing a view of a wall of an object relatively mobile relative to said capturing device. To highlight a surface defect in such a wall, such a device comprises:

- a matrix sensor oriented so that an axis of the capture field is substantially normal to the wall of the object;

- A first light element comprising a first light source, said first light element being arranged to project a first determined and heterogeneous pattern on the wall of the object when the latter is movable relative to the capture device;

- The matrix sensor and said first light element being mutually positioned so that at least one light ray emitted by the first light source is reflected by the wall of the object and captured by the matrix sensor.

To constitute the first determined and heterogeneous pattern, a first embodiment of a capture device according to the invention may consist in that the first light element comprises a mask arranged to partially obscure the light rays delivered by the first light source, the first determined and heterogeneous pattern consisting of a plurality of light bands.

Alternatively, the first light source of the first light element of a capture device according to the invention may consist of a screen diffusing an image describing a plurality of light bands.

To provide assistance in detecting a change in tint in the wall of the object, a capture device according to the invention may comprise a second light element comprising a second light source, said second light element being arranged to project a second determined and homogeneous pattern on the wall of the object when the latter is movable relative to the capture device, the matrix sensor and said second light element being mutually positioned so that at least one light ray emitted by the second source light is reflected by the wall of the object and captured by the matrix sensor, the second light source being chosen to project the homogeneous pattern according to a light intensity lower than that of the first pattern projected by the first light element.

To reduce the time necessary for the total capture of the wall of the object during the entirety of the relative displacement of the object with regard to a device for capturing views representation produced in the last device can partially of the capturable field capture, increase the frequency of while preserving the final fine digital by capture, deliver the definition of the said one the one integrating a matrix wall such a sensor system digital representation by said sensor, said of this partial digital representation comprising a determined number of columns of pixels .

In order to be able to be integrated into a system for assisting in the detection of a surface defect in the wall of the object, the invention provides that the matrix sensor and the first light source can be arranged to be controlled electrically, respectively for capture a view and to emit light rays, by a processing unit cooperating with said matrix sensor and said first light source according to a wired or wireless link.

In this case, according to the configuration capabilities offered by the matrix sensor of a capture device according to the invention, said matrix sensor can be arranged to size the partial digital representation according to the value of a configuration parameter delivered by the processing unit.

According to a second object, the invention also relates to a system for assisting in the detection of a surface defect on a wall of an object comprising:

a capture device according to the invention, as previously expressed;

a processing unit cooperating in a matrix and / or or wireless wire connection with the sensor the first light source of the first light element of said capture device;

said processing unit being arranged for:

o iteratively trigger, according to a matrix sensor frequency capture and / or determined, the first light source of the first light element of said capture device and receive from said matrix sensor a digital representation of a view captured by said matrix sensor;

o juxtapose the digital representations delivered by the matrix sensor according to a chronological order of capture and obtain a composite and / or merged representation describing a first wall of the object.

To constitute a history of captures of an object, the processing unit of such a system can comprise, or cooperate with, a data memory, said processing unit being arranged to write into said data memory at least one representation digital received and / or produced by said processing unit.

When the matrix sensor of a capture device of a help system according to the invention cannot constitute a partial representation of sufficiently reduced size, the processing unit of said system can be arranged to produce a partial digital representation resulting from the digital representation delivered by said sensor, said partial digital representation comprising a determined number of columns of pixels.

To constitute a bank of digital representations describing the wall of an object to an entity distant from said detection assistance system, the processing unit of said system can include communication means ensuring a link with a remote storage server arranged to receive and store at least one digital representation produced by the processing unit and transmitted by said communication means.

To allow an accelerated audit of the object and thus reduce the number of passages of said object in front of a matrix sensor, a system according to the invention may include a second capture device also in accordance with the invention, said first and second capture devices being mutually arranged so that the axis of the capture field of the matrix sensor of said second capture device is

oriented substantially normal to a second wall of the object when the latter describes a relative movement with respect to said first and second capture devices.

According to an advantageous and economical embodiment, the processing unit of a system for assisting in the detection of a surface defect according to the invention can jointly process the data delivered by the first and second capture devices. For this, the invention provides that:

- the processing unit can cooperate in a wired or wireless connection with the matrix sensor and / or the first light source of the first light element of said second capture device; said processing unit can be arranged for: o trigger iteratively, according to a determined capture frequency, the matrix sensor and / or the first source light of the first light element of said second capture device and receiving from said matrix sensor a digital representation of a view captured by said matrix sensor; o juxtapose the digital representations delivered by the matrix sensor in a chronological order of capture and produce a composite representation and / or merged describing the second wall of the object and.

According to a preferred application, such a mutual arrangement of the first and second capture devices of a system according to the invention can ensure the capture of respective views of two distinct sections of a bodywork of a land vehicle.

send to

1 state of an ob j and paro before capturing a defect such

For easy passage of said detection device 1 invention, representation interface the view of said system automatically representation of said human digital man-machine unit.

ob j and sensor system surface can to make the human unit to analyze at the end of a matrix of an aid to the conform to include a display a perceptible by treatment to cause graph of a produced by operator i of a one of a suitable system and the

In this case, a composite restitution and / or treatment can be arranged.

fused

Other characteristics and advantages will appear more clearly on reading the following description and on examining the accompanying figures, among which:

- Figure 1 shows a simplified view of a capture device according to the invention;

- Figure 2 illustrates a preferred but non-limiting example of a heterogeneous pattern comprising a plurality of light lines that can project on one of the walls of an object a first light element of a capture device according to the invention ;

- Figure 3 describes an example of capturing a surface defect that can be captured by a capture device according to the invention;

- Figure 4 describes a functional architecture of a system for assisting in the detection of a surface defect according to the invention.

FIG. 1 describes a simplified view of a nonlimiting example of a capture device 1 of a system for assisting in the detection of a surface defect in a PLV wall of an object, in this case , the body of a vehicle

V, passenger side. According to the embodiment described in connection with FIG. 1, said vehicle

V performs a displacement

Linear and sensitive DPL normal to the capture axis

AC of a matrix sensor

C of said capture device 1. Such a matrix sensor C can consist of one or more cameras such as, by way of nonlimiting example, a camera

Monochrome Firefly MV from the manufacturer FLIR which can be connected to a processing unit not shown in Figure 1.

Such a connection can consist of a wired link such as a FireWire bus (trade name assigned to a multiplexed serial interface describing a computer bus carrying both data and control signals) or a USB bus (Universal Serial Bus according to English terminology) -saxonne) describing a computer bus in serial transmission. Any other camera could, as a variant, be used instead or in addition. In the same way, the cooperation between the matrix sensor C and such a processing unit could alternatively consist of a wireless link according to different communication protocols such as for example Bluetooth, Wi-Fi or ZigBee. By way of nonlimiting example, the matrix sensor C can deliver a digital representation of 752 by 480 pixels, each time a view is captured, if said sensor is a monochrome Firefly MV type camera.

The side wall PLV of the vehicle V that said matrix sensor C can capture roughly fits in a relatively mobile plane with respect to the capture axis AC of said sensor. Depending on the dimensions of the wall of said vehicle whose surface is to be analyzed, the matrix sensor C can consist of a set of several sensors or cameras, each capturing a section of said wall.

In order to be able to detect a surface defect linked to a deformation of said surface of the wall of the object V, in this case a vehicle, such as for example the deformation of a wing of an automobile following an impact , a capture device 1 according to the invention comprises a first light element EL1 comprising a first light production source or first light source L1. Such a light source L1 can consist of one or more computer or television screens, the transmission power of the light signal which it (s) projects (s) can optionally be manually or electronically adjusted. As a variant or in addition, such a first light source L1 may consist of one or more light producing elements such as, for example, light-emitting diodes, or even halogen incandescent lamps, although they are more energy-consuming than said diodes. To facilitate the detection of deformations of the surface of the PLV wall of the object V, it is particularly advantageous for the first light element EL1 to project a determined and heterogeneous pattern, for example a pattern consisting of a plurality of vertical light lines, or more generally, normal to the surface on which the object V moves DPL and / or, as a variant or in addition, normal to the plane in which the matrix sensor C moves if the object remains stationary. When the first light source L1 consists of a computer and / or television screen, the projected pattern can be defined such as a test pattern as shown in FIG. 2. Such a first determined and heterogeneous pattern can thus consist of a plurality of CW light lines or bands, for example of a white color, said CW light lines or bands being separated in pairs by a black band or line CB. Such a first pattern is said to be heterogeneous because it does not consist of a uniform pattern, which we could qualify as homogeneous because identical in any point of emission of the first light element EL1.

When said first light source L1 consists of one or more traditional bulbs or lamps, the light rays which said first light source L1 can project can be partly obscured by a PT mask allowing only certain light rays to pass through, thus forming a test pattern similar to that mentioned above when said first light source L1 consists of a computer screen. Other forms of first light sources L1 could be used so as to form a first determined and heterogeneous pattern which can be projected onto the wall PLV of the object V during its relative displacement DPL.

FIGS. 1 and 3 also make it possible to image the principle making it possible to highlight a deformation of the PLV surface of an object V. As indicated in FIG. 1, the first light element EL1 and the matrix sensor C are mutually arranged so that a light ray RI, emitted by the first light source L1 and projected onto the wall PLV, substantially distant from said first light source L1 by a distance D, can be reflected at a point PI of said wall PLV, then captured in the form of a reflected light ray RI 'by the matrix sensor C. Said reflected ray RI' follows a linear trajectory from the point PI deviated by an angle Θ with respect to a normal NI to the wall PLV at said point PI. Thus, during the relative DPL movement of the object V relative to the matrix sensor C, any significant variation in the angle of reflection Θ reflects a deformation of the surface of the wall PLV. If the light ray RI is emitted from a row r of the first light source L1 (“row” means the abscissa of an emission point of a ray along an axis, inscribed in the surface d 'emission of the first light source L1, parallel to a plane in which the relative displacement of the object V is registered relative to said first light source L1), there are 7 * 0 the relation - = 2Dsin-. Thus, a hollow deformation of the surface of the PLV wall results in a significant modification of the angle of reflection Θ of a radius RI emitted from the same row r in a plane substantially parallel to the surface on which DPL l 'moves object V and / or the matrix sensor C. Such a reflection angle Θ would in this case increase strongly according to the depth of the surface deformation and then recover a value substantially equal to that outside of any surface deformation, then decreasing to become negative before recovering said value of the angle Θ in the absence of any deformation of the surface of the PLV wall. The reflection of the first heterogeneous pattern which can be captured by the matrix sensor C is therefore advantageously indicative of a deformation of the PLV surface.

FIG. 3 thus describes a situation in which an object V, for example an automobile, moves DPL in front of a matrix sensor C projecting a first determined and heterogeneous pattern thanks to a mask PT partially obscuring the light rays emitted by the first source light Ll of the first light element EL1 of a capture device according to the invention. According to this example, one of the side doors of the automobile V was damaged and a defect DF forming a hollow on the surface of the wall PLV of said door exists. The first heterogeneous pattern describes a series of n light lines CW1, CW2, CWi to

CWn vertical, the automobile V describing a horizontal movement DPL on the ground. The matrix sensor C can capture one or more lateral views of said automobile V and thus produce, on each capture, a digital representation R in the form of a table of pixels of RL rows and RC columns. As shown in Figure 3, we denote P x, y, the pixel at the RRx ^th line and CCyR ^th column. As shown in FIG. 3, certain matrix sensors can produce one, or even several, partial digital representations RW comprising, for example, only certain columns of pixels and / or certain lines of the representation R which embraces the whole field captured by said sensor. The dimensioning of a partial representation can be, for example, predetermined or configurable by electrical control. It is the same for a line or column offset, in order to precisely position the pixels selected to produce such a partial representation RW. By way of example, FIG. 4 describes a sensor capable of producing a partial representation of RWw columns of pixels. Whatever the representation produced, whole R or partial RW, the captured view is that described by the example of the passenger door of a damaged automobile presented in FIG. 3. Thus, in said FIG. reflected light CWi 'to CWn', respectively originating from the lines CWi to CWn of said first heterogeneous pattern projected by the first light element EL1, said reflected light bands not being deformed by the wall PLV of the vehicle. Conversely, we can perfectly distinguish that the CWI 'and CW2' light bands, reflected by the PLV wall, respectively from the CWI and CW2 light bands of the first heterogeneous pattern projected by EL1, are deformed, even broken, by the presence of a defect. DF of said PLV wall surface.

FIG. 1 describes a particularly advantageous mutual arrangement of the first light element EL1 and of the matrix sensor C of a capture device according to the invention. Thus, if a normal to the surface of emission of the light ray or rays RI by the first light source L1 describes an angle substantially of 45 ° with the capture axis AC of the matrix sensor, the reflection of the light rays forming the first heterogeneous pattern occupies only a first part of the field which can be captured by said matrix sensor C, leaving a second part of said field which can be captured by said matrix sensor C not affected by such a reflection.

Such a second part of said capture field can be used to capture other light rays also reflected by the wall PLV of the object V during the relative DPL movement of the object V with respect to said sensor. Thus, in accordance with a preferred arrangement and not described in Figure 1, a capture device according to the invention may include a second light element EL2.

The latter comprises a second light source L2, however chosen to emit light rays

R2 of an intensity lower than that of the ray (s) RI corresponding to the first heterogeneous pattern. Said second light element EL2 is arranged to project a homogeneous pattern, that is to say that the light intensity of any ray emitted at any point on the emitting surface of said second light element EL2 is substantially constant. Thus, the arrangement of the second light element EL2 makes it possible to project a second determined and homogeneous pattern on the wall PLV of the object V. Like the first light element

EL1, the second light element EL2 and the matrix sensor C are mutually arranged so that a ray

R2 emitted by the second light source L2, reflected at a point

P2 of said PLV surface in the form of a reflected ray R2 ′, can be captured by said matrix sensor C.

The arrangement of a capture device described by the figure teaches that the first and second light elements EL1 and

EL2 can be positioned respectively on either side of said matrix sensor C. According to this advantageous but non-limiting arrangement, a normal to the emission surface of the second determined and homogeneous pattern of said second light element EL2 forms an angle substantially equal to

45 ° with the capture axis of said matrix sensor C.

Thus, the reflection of the second homogeneous pattern may occupy, depending on the field opening proposed by the matrix sensor C, only a part of the field which can be captured by said matrix sensor

C, said part being distinct from that concerned by the reflection of the first heterogeneous pattern. In addition, the light intensity of said second homogeneous pattern being advantageously chosen to be less than the light intensity of the light ray (s) RI composing the first heterogeneous pattern, the conjunction of the two patterns has no impact for capturing the reflection of said first heterogeneous pattern. Thus, with one and the same matrix sensor C, it is possible, in a single view capture, to collect information related to the respective reflections of the first and second determined patterns. As long as we only consider a partial capture or restricted region of interest RW in the digital representation

R describing the entire field that can be captured by said matrix sensor C, it is possible to take into account only the reflection of the first or second determined pattern, as we will detail below in connection with the example of implementation of '' a system to help detect a surface defect illustrated in Figure 4.

The second light source L2 can consist, like the first light source L1, of a computer or television screen or a network of light-emitting diodes. It can also be summed up as a passive screen, such as a cinema screen, in a clear or white color reflecting the ambient light of the object evenly.

Let us focus more specifically on the implementation of a capture device by a system for assisting in the detection of a defect in a surface of a wall of an object. Such a situation is illustrated in particular in FIG. 4 by way of example of nonlimiting application. According to this example, such a system includes a capture device 1 according to the invention and as described in connection with FIG. 1.

According to a first embodiment, said capture device 1 comprises a matrix sensor C and a first light element EL1. The first light source L1 of said first light element EL1 and the matrix sensor C of the capture device 1 cooperate with a processing unit 11, comprising one or more microprocessors or microcontrollers, such a processing unit 11 being arranged to implement a method in particular for controlling the two materials, said implementation possibly being caused by the execution of instructions from a suitable computer program loaded into a program memory 13 cooperating with said processing unit or cooperating with it. Such a processing unit 11 can be that of a personal computer or, as a variant, of a dedicated control unit. As mentioned above in connection with FIG. 1, the matrix sensor C and / or the first light source L1 of said capture device 1 can cooperate in a wired or wireless connection with the processing unit 11. According to the example described by FIG. 4, the latter is arranged to configure (situation symbolized by a link referenced Pc in FIG. 4) the matrix sensor C, in order to define in particular a selection of pixels of interest constituting partial digital representations RW of the capture field of said sensor matrix C. Exploiting such restricted or partial representations RW instead of digital representations R describing the entirety of said capture field of the matrix sensor C makes it possible in particular not to burden the image processing carried out by the processing unit 11 between two screenshots and thus not to penalize the speed of the relative movement DPL of the object V compared to the d capture device 1.

The processing unit 11 of a system according to the invention can be arranged to activate the first light source L1 when an object is about to enter the capture field of the matrix sensor C. For this, the unit 11 can cooperate with a presence sensor, not shown in Figure 4 for simplification, said sensor detecting the presence of an object which is sought to analyze the envelope. Such a presence sensor can consist of an optical or pneumatic presence detection sensor, like tubular sensors used on tracks to detect the passage of a vehicle, or even an inductive sensor. Said processing unit 11 can also be arranged to cause or control electrically, or even by electromagnetic waves, the capture of a view by the matrix sensor C.

When such an object V enters an analysis space of one of its PLV envelope walls, the processing unit 11 of a system according to the invention iteratively causes a series of captures, carried out by the sensor matrix C, during the relative displacement of the object V, as in the present case in connection with FIG. 4 an automobile. Said matrix sensor C thus produces a series of partial representations RW (t) to RW (tm) capturing the reflection of the first heterogeneous pattern on the wall PLV of the mobile object V. The partial representations RW (t) to RW (tm) can be written to a data memory 12 specific to the processing unit 11 or cooperating with the latter via a data bus. At the end of the passage of the object V, the processing unit 11 causes the implementation of a joint processing of the different partial representations RW (t) to RW (tm) to produce by aggregation a composite digital representation RX describing the entire PLV wall of vehicle V.

At the end of the production of a composite RX representation, the processing unit 11 can implement a step of merging the partial representations aggregated in the RX representation to produce a unified digital representation RF of said wall PLV of the object V in the form of an image. This production can be the subject of a recording of the RX or RF representations in the data memory 12. The processing unit 11 can also cause a graphical restitution of said RF image by a human-machine interface DP such as a computer screen for example or by any other suitable output device, such as a printer. An operator can thus instantly take cognizance of the RF image and check the presence of a possible DF surface defect highlighted by the reflection of the first heterogeneous pattern. According to the example described in connection with the figure

4, such an RF image makes it possible to facilitate

The operator's analysis and thus allow him to detect the presence of a surface defect on the PLV side wall.

The invention also provides that the processing unit 11 of a compliant system can configure the matrix sensor ci jointly delivers, during a to one invention so that the capture of the current scene, two separate representations RW (t), each one reflections generally focusing thus on points of distinct on said variant wall, can be can produce if

PLV, at the PLV wall, or more distinct regions of interest within the capture field.

of

In a single product, causing said partial representation RW does not capture, the processing unit two two consecutive captures for separate partial representations RW (t). Thus, finally, two RF images can be produced by the processing unit 11 on the basis of the two series of partial representations

RW (tm) devoted to the two regions of interest.

The operator can thus order a joint or concomitant display, even alternately, of the two images

RF, each giving in a way two images virtually taken from different angles of view, highlighting more a surface defect, the reflection of the first heterogeneous pattern giving the operator the impression of a displacement on the surface of the POS wall. As a variant, such a display can be automatically caused by said processing unit.

In the same way as for the production of an RF image resulting from the fusion of partial representations RW reflecting the reflection of the first heterogeneous pattern projected by the first light element EL1 of the capture device 1, the invention provides that the unit of processing 11 of a system for detecting a surface defect can produce a digital RX and / or RF representation of the POS wall lit by a second determined pattern, this time homogeneous, offering the operator a aids in the detection of surface surface defects such as a scratch, a scratch or a change in color on said PLV wall of the analyzed object V. For this, the capture device 1 of said system comprises a second light element EL2 arranged to project onto the wall of the object a second homogeneous pattern, as already explained in connection with FIG. 1.

In the case where the capture field of the matrix sensor C is wide and allows, on the one hand, to collect the reflection on the envelope or PLV wall of the object V of the first heterogeneous pattern projected by the first light element EL1 and on the other hand, the reflection on said envelope or PLV wall of the second homogeneous pattern projected by the second light element EL2, without the two reflections completely overlapping, the processing unit 11 can cause a capture of the scene and control to said matrix sensor C the concomitant production of a first partial digital representation RW (t) capturing all or part of the light rays coming from the first heterogeneous pattern, and a second partial digital representation RW (t), capturing all or part of the light rays coming from of the only second homogeneous motif.

When the matrix sensor C can only capture a view according to which the reflections of the first and second patterns, respectively heterogeneous and homogeneous, overlap, the invention provides that the processing unit 11 can cause, in place of a single capture and production of two partial representations RW (t), a first capture, the first light source L1 being actuated by the matrix sensor, the latter producing a representation describing a reflection of the first heterogeneous pattern on the PLV wall, followed by a second capture, the first light source L1 being extinguished under the action of a command transmitted by said processing unit

11, by said matrix sensor, the latter producing a second digital representation only by the reflection of the second homogeneous pattern on said wall

POS. Since the intensity of emission by the second light element EL2 of the second homogeneous pattern can be advantageously chosen to be lower than the intensity of emission by the first light element EL1 of the first heterogeneous pattern, the possibly permanent projection of the second homogeneous pattern does not not capture the reflection of the first heterogeneous pattern on the wall

POS.

An alternative display of two RF images, one describing the reflection on the PLV wall of the first heterogeneous pattern and the second describing the reflection of the homogeneous pattern on said PLV wall also offers the operator valuable assistance in detecting a DF surface defect.

Depending on the technology of the matrix sensor C, it is possible that the latter cannot directly produce a partial representation RW (t) describing a scene of which the unit captures. In this case, the invention provides for processing 11 to be responsible, upon receipt of a digital representation R (t) describing the entire capture field, for producing one or the digital representation R (t), instead and place of said matrix sensor C as mentioned above.

The invention also provides that the processing unit 11 may include means, not shown in FIG. 4, for time stamping the production of a composite digital RX representation or of an RF image. It is the same for the matrix sensor C. This can also be arranged to time stamp any digital representation that it produces and delivers to the processing unit 11. Such a time stamp can be advantageously associated, entered or even embedded in any digital representation produced according to the invention. Furthermore, the invention provides that a system for assisting in the detection of a surface defect may include other sensors or entities cooperating with the processing unit 11 to identify an object V in the process of being captured during of his passage DPL. Such an additional or complementary sensor can be optical to capture, for example, the content of a front or rear registration plate of a vehicle V, or even electromagnetic, to read an identification message emitted by a beacon present on or in object V, or of any other nature. The processing unit 11 can then associate additional information delivered by such additional sensors with the digital RX or RF representations produced, or even embed alphanumeric characters within said representations describing this additional information.

Furthermore, a processing for analyzing the content of RX, RF images or representations produced by a system for assisting in the detection of a surface defect in accordance with the invention can be implemented by a remote electronic entity. , such as an audit or storage server. As a variant or in addition, all or parts of said RX or RF digital representations, or even R (t) or RW (t) representations produced by the matrix sensor C or by the processing unit 11, can be the subject of a recording on such a remote audit server 20. For this, the invention provides that the processing unit 11 of such a system can include or cooperate with communication means 14 ensuring a NW connection, for example of the type Internet or GPRS (General Packet Radio Service according to Anglo-Saxon terminology), or even satellite, with said remote electronic entity 20. The processing unit 11 is then arranged to produce data messages encoding such digital representations and to cause the transmission of such data messages by said communication means 14 to said remote electronic entity 20.

FIG. 4 made it possible to illustrate different embodiments of a system for assisting in the detection of a surface defect in a PLV wall of an object V, such as an automobile.

In order to be able to analyze several PLV walls of said object during a single relative DPL movement of the latter, a system in accordance with the invention may include a plurality of capture devices also in accordance with the invention, as described by way of example by the figure 1.

Thus, by way of nonlimiting example, such a system may include a second capture device 1 ′, similar to the capture device 1 described in connection with FIGS. 1 and 4, the second device 1 ′ facing the device for capture 1. Thus, the two capture devices 1 and 1 ′ can deliver to the processing unit 11 representations R (t) or RW (t) respectively describing the two side walls of the automobile V illustrated in FIG. 4 In this configuration, the first and second capture devices 1 and 1 'are mutually arranged so that the axis AC of the capture field of the matrix sensor C of said second capture device 1' is oriented substantially normal to the wall of the bodywork, driver's side of the automobile V when the latter describes a relative displacement DPL with respect to said first and second capture devices 1, 1 ′, the axis AC of the capture field of the matrix sensor C of the first device tif capture 1 being meanwhile substantially normal to the wall of the body of said automobile, passenger side.

The invention further provides that a third capture device can in turn capture, for example, views from above of vehicle V. The three capture devices thus form the equivalent of a bridge within which a land vehicle can move, or vice versa, a movable bridge moving along a longitudinal axis of said vehicle V.

When a system according to the invention comprises a plurality of capture devices, such as the devices 1 and 1 ′ mentioned above, the processing unit 11 of said system can cooperate in a wired or wireless connection with the matrix sensor C and / or the first light source L1 of the first light element EL1 of each of said capture devices. It is then also arranged for:

- iteratively trigger, according to a determined capture frequency and adapted to the authorized speed of movement of the vehicle V during its DPL pass, the matrix sensor C and / or the first light source L1 of the first light element EL1 of each capture device and receiving or receive from said matrix sensor a digital representation R (t) or RW (t) of a view or scene captured by said matrix sensor C of each capture device;

- juxtapose the partial digital representations thus delivered by the various matrix sensors C according to a chronological order of capture and obtain a composite RX and / or merged RF representation describing the various PLV walls of the object V.

When such a system for detecting a surface defect in accordance with the invention is installed, for example in a land vehicle rental center, it therefore becomes possible to easily and objectively check the state of the bodywork of a vehicle when the said vehicle is made available to a tenant or when the said vehicle is returned by the latter.

The digital RX, RF representations produced in accordance with the invention by a processing unit of a system for assisting in the detection of a surface defect, are also liable to be the subject of a subsequent automatic processing of implementation. such flaws emerge, to further facilitate and maximize the objectivity of such an examination. Such additional processing could be implemented by the processing unit 11 of a system according to the invention or by an electronic entity 20 remote from it as illustrated in FIG. 4.

The invention has in particular been described in connection with an application for analyzing a bodywork and / or a chassis of a land vehicle. However, it could be applied to any vehicle for which the integrity of the outer envelope is to be checked and even to any other object. It will suffice simply to adapt, size and arrange in particular the first and second light elements EL1 and / or EL2 of the capture device (s), with regard to the matrix sensors used to analyze the surface of one or more walls of the object concerned .

Other modifications can be envisaged without departing from the scope of the present invention defined by the appended claims.

Claims (15)

1. Device (1) for capturing a view of a wall (POS) of an object (V) relatively mobile relative to said capture device (1), the latter (1) being characterized in that it includes: - a matrix sensor (C) oriented so that an axis of the capture field (AC) is substantially normal to the wall (POS) of the object (V); a first light element (EL1) comprising a first light source (L1), said first light element (EL1) being arranged to project a first determined and heterogeneous pattern on the wall (POS) of the object (V) when it it is movable relative to the capture device (1); - the matrix sensor (C) and said first light element (EL1) being mutually positioned so that at least one light ray (RI) emitted by the first light source (Ll) is reflected (PI) by the wall (PLV) of the object (V) and captured (RI ') by the matrix sensor (C). 2. Device (10) according to the preceding claim, for which the first light element (EL1) comprises a mask (PT) arranged to partially obscure light rays (RI) delivered by the first light source (Ll) and constitute the first pattern determined and heterogeneous, so that the latter consists of a plurality of light bands (CW). 3. Device (10) according to claim 1, for which the first light source (L1) consists of a screen diffusing an image describing a plurality of light bands (CW). 4. Device (10) according to any one of the preceding claims, comprising a second light element (EL2) comprising a second light source (L2), said second light element (EL2) being arranged to project a second determined and homogeneous pattern on the wall (PLV) of the object (V) when the latter is movable relative to the capture device (1), the matrix sensor (C) and said second light element (EL2) being mutually positioned, so that at less a light ray (R2) emitted by the second light source (L2) is reflected (P2) by the wall (PLV) of the object (V) and captured (R2 ') by the matrix sensor (C), the second light source (L2) being chosen to project the homoqene pattern according to a light intensity lower than that of the first pattern projected by the first light element (EL1). 5. Device according to any one of the preceding claims, for which the matrix sensor (C) delivers a partial digital representation (RW) of the field which can be captured by said sensor (C), said partial digital representation (RW) comprising a number (RWw ) of determined pixel columns. 6. Device according to any one of the preceding claims, in which the matrix sensor (C) and the first light source (L1) are arranged to be electrically controlled, respectively to capture a view and to emit light rays (RI), by a unit processing (11) cooperating with said matrix sensor (C) and said first light source (L1) in a wired or wireless connection. 7. Device according to claims 5 and 6, for which the matrix sensor (C) is arranged to size the partial digital representation (RW) according to the value of a configuration parameter (Pc) delivered by the processing unit (11 ). 8. System for assisting in the detection of a surface defect (DF) on a wall (PLV) of an object (V), characterized in that it comprises: - a capture device (1) according to claims 6 or 7; - a processing unit (11) cooperating in a wired or wireless connection with the matrix sensor (C) and / or the first light source (L1) of the first light element (EL1) of said capture device (D; - said processing unit (11) being arranged for: o iteratively trigger, according to a determined capture frequency, the matrix sensor (C) and / or the first light source (Ll) of the first light element (EL1) of said capture device (1) and receive from said matrix sensor a representation digital (R (t ), RW (t) , RW (tm)) of a captured view through said sensor matrix (VS) , o juxtapose the representations digital ( R (t), RW (t), RW (tm)) issued through the matrix sensor (C) according to an order chronological capture and produce a representation composite (Rx: ) and or merged (RF) describing a first wall (POS) of The object (V). 9. System according to the preceding claim, for which the processing unit (11) comprises, or cooperates with, a data memory (12), said processing unit (11) being arranged to write into said data memory (12 ) at least one digital representation received (R (t), RW (t), RW (tm)) and / or produced (RX, RF) by said processing unit (11). 10. System according to claims 8 or 9, for which the processing unit (11) is arranged to produce a partial digital representation (RW) resulting from the digital representation (R) delivered by said sensor (C), said digital representation partial (RW) comprising a determined number (RWw) of columns of pixels. 11. System according to any one of claims 7 to 9, for which the processing unit (11) comprises or cooperates with communication means (14) ensuring a connection (NW) with a remote electronic entity (20) arranged to receive at least one digital representation produced (RC, RF) by the processing unit (11) and transmitted by said communication means (14). 12. System according to any one of a second claims comprising (1 ') conforming to first d, 1') 1 axis and device of claims 6 or second devices arranged capture substantially normal device being mutually (AC) of the field of (C) of said second 7, of the capture so that the matrix capture sensor (1 ') is oriented to a second wall (PLV) thereof describes a relative displacement (DPL) with respect to said first and second capture devices (1, 1' ). 13. System according to the preceding claim, for which: - the processing unit (11) cooperates in a wired or wireless connection with the matrix sensor (C) and / or the first light source (L1) of the first light element (EL1) of said second capture device); - said processing unit (11) being arranged for: o iteratively trigger, according to a determined capture frequency, the matrix sensor (C) and / or the first light source (L1) of the first light element (EL1) of said second capture device (1 ') and receive a representation of said matrix sensor digital (R (t), RW (t ), RW (tm)) of a captured view by said matrix sensor (VS) ; o juxtapose the representations digital (R (t), RW (t), RW (tm)) delivered by the matrix sensor (C) in a chronological order of capture and produce a composite (RX) and / or merged (RF) representation describing the second wall (PLV) of the object (V). 14. System according to claims 12 or 13, for which the mutual arrangement of the first and second capture devices (1, 1 ') is determined to capture respective views of two separate sections (PLV) of a bodywork of a land vehicle (V). 15. System according to any one of the claims 8 to 14, comprising a man-machine interface (DP) adapted to display a digital representation and make it perceptible by the sight of a human and for which the processing unit is arranged to cause a graphic reproduction of a composite representation (RX) and / or merged (RF) produced by said processing unit (11). 1/2