WO2023118953A1

WO2023118953A1 - Device and procedures for human-like automatic inspection of defects

Info

Publication number: WO2023118953A1
Application number: PCT/IB2021/062335
Authority: WO
Inventors: Boris BRET; Helder CORREIA; André AMORIM; José M. GONZÁLEZ MÉIJOME
Original assignee: Bosch Car Multimedia Portugal, S.A.; Universidade Do Minho
Priority date: 2021-12-22
Filing date: 2021-12-27
Publication date: 2023-06-29

Abstract

The present application describes an apparatus and a method to perform human-like automatic inspection of defects in objects. The proposed optomechanical system for automatic inspection of defects in objects, comprises a camera casing comprising an optical sensor located inside said casing; a posterior segment of tubular hollow shape mechanically adjusted to said camera casing through thread means comprised in an outer surface of the posterior segment and comprised in an inner surface of an existing circular opening located in a side face of said casing; and an anterior segment of tubular shape mechanically adjusted to said posterior segment through thread means comprised in an outer surface the anterior segment and comprised in an inner surface of the posterior segment. The system is configured to reproduce the human eye inspection by mimicking its optical characteristics, variable resolution adjustment and tunable distance inspection adjustment.

Description

DESCRIPTION "Device and procedures for human-like automatic inspection of defects"

Technical Field

The present application describes an apparatus and a method to perform human-like automatic inspection of defects in ob j ects .

Background art

Document US 6485142B1 describes an arti ficial eye that includes a generally spherically shaped container including a substantially hemispherical posterior portion, a substantially hemispherical anterior portion and a fastener for attaching the posterior portion to the anterior portion . The arti ficial eye includes functional counterparts to the anterior and posterior chambers of the human eye . Within these chambers are fluids which mimic the characteristics of the aqueous and vitreous humors in the human eye . Protective eyewear is tested by placing the eyewear between the arti ficial eye and a source of radiation . Radiation is directed through the protective eyewear and then through the arti ficial eye . The radiation is then received by a sensing device such as a charged couple device ( CCD) camera or optometer . The resulting image is then evaluated .

Document US7066598B2 discloses an eye model that includes a partially-spherical transparent disk for representing a cornea, and a housing attached to the disk including a volume for holding a fluid representing aqueous within an eye . An annular ring is attached to the housing for representing an iris and a lens is attached to the annular ring for representing a lens of the eye . An end-cap on an end opposite the disk 12 represents a retina .

Summary

The present invention describes optomechanical system for automatic inspection of defects in obj ects , comprising a camera casing comprising an optical sensor located inside said casing; a posterior segment of tubular hollow shape mechanically adj usted to said camera casing through thread means comprised in an outer surface of the posterior segment and comprised in an inner surface of an existing circular opening located in a side face of said casing; and an anterior segment of tubular shape mechanically adj usted to said posterior segment through thread means comprised in an outer surface the anterior segment and comprised in an inner surface of the posterior segment ; wherein the thread means of the mechanical adj ustment between the anterior segment and the posterior segment comprise a predetermined length that allows to fine tune the distance between an anterior extremity of the anterior segment and anterior extremity of the posterior segment .

In a proposed embodiment of present invention, the optical sensor is configured to be set at a predetermined distance from a posterior extremity of the posterior segment accordingly to camera lens mounting standards , said distance defining the flange focal distance .

Yet in another proposed embodiment of present invention, the anterior segment comprises a groove in the anterior extremity to allow the mechanical fitting of a holder which traverses a central portion of the tubular hollow shape .

Yet in another proposed embodiment of present invention, the anterior segment comprises a tubular hollow shaped angular element mechanically fitted between the anterior extremity and the thread means the anterior segment .

Yet in another proposed embodiment of present invention, the holder comprises an optoelectronic lens arrangement with fixed and/or variable apertures , adj usted through distancing of anterior and/or posterior segments separation .

Yet in another proposed embodiment of present invention, the fine tune of the distance between an anterior extremity of the anterior segment and anterior extremity of the posterior segment and the optoelectronic lens arrangement with variable apertures allows to capture and reproduce images of extended obj ects on a wide field o f view within on-axis and of f-axis and changing a distance of observation adj usting the region of interest (ROI ) in case of a detection of non- obvious defects .

The present invention further describes the method of operation of the optomechanical system for automatic inspection of defects in obj ects , according to any the previous description, comprising the steps of System initiali zation; Position reset of the acquired image to full si ze range ; Acquire image ; Analyse acquired image data ; Categori zing defect i f a defect found in the analysed acquired image data, recategori zing said defect in case of recurrence through adj ustment of the detail and region of interest of the acquired image ; Production of the final report .

In a proposed embodiment of the method of operation of the optomechanical system for automatic inspection of defects in obj ects , the categori zing defect comprises additional searching for defects in the acquired images data through adj ustment of the position of the optomechanical system with regard to the obj ect .

General Description

The present invention describes an optomechanical eye model ( OME ) that applies the optical and resolution characteristics of the human eye to automatically detect potential defects in display systems , automotive parts or any other obj ect where such defects can weaken the quality of the product by being seen by the human user .

The inspection procedure for defects and detection in automotive industry comprises tight tolerances and requirements . Some speci fic production parts are critical , as in the case of displays , because they are under the scrutiny of users expecting high quality standards in the vehicles they acquire . To ensure reduced return rates and the related production and reputational costs , industry invest in very sensitive inspection methods , that ultimately pushed the standards to limits that might not have an impact on the user experience i f they are not seen under the limits of resolution of the human eye . Compared to other physiological systems , there is a strong interest in physical modelling the optics of the eye . Due to their small dimensionality and complexity, mere computer theoretical models usually lack enough precision to assess the quality of vision, design optical implants or plan and monitor treatments . One of the main challenges is to be able to predict the subj ective response of a patient prior to actually altering the patient ' s ocular characteristics , either surgically or by external means such as spectacles or contact lenses . An additional problem is the fact that the eye characteristics present large inter-individual variability, which means that the models need to be highly customi zable . Thus , the ability to accurately predict the in-vivo performance of a constantly changing ocular parameters has many obvious advantages over theoretical computer models , with the goal of reducing the time , cost , and uncertainty involved in ef forts to improve visual quality . Applied to optical inspection, physical optomechanical models of the eye might be able to replicate the human eye conditions and emulate the human inspection providing consistency, adaptabil ity and scalabil ity to the quality control process .

Various ef forts had been made to build realistic physical eye models . Nowadays , eye models are way more complex, detailed and varied, being widely used as benchmarks to evaluate the performance of spectacles , contact lens and intraocular lens ( IOL ) , being also used to test and calibrate optical measurement devices , and for use in research studies .

Most physical eye models are usually either mounted on a workbench or built as an independent single device . In the former case , the parts are assembled together without the need to be included in a single unit . This has the inconvenience that the system needs to be mounted on a table , occupying a signi ficant amount of space , and not being easily moved around . On the other hand, these models can be more complex, highly configurable on the workbench and integrated with other hardware , and therefore they can be set up to fit very speci fic needs . As for independent single devices , most parts are usually attached to the same enclosure . These models are much smaller and more portable , easy to mount , and also resemble more a real eye . The enclosure is usually spherical as a real eye , despite this does not af fect in any way the formation of images and is done merely for aesthetics . Usually, these devices comprise of two halves that are attached together for easier mounting . Unfortunately, these models usually include less features and are less configurable , and in many cases the parts are glued or cemented together, meaning they cannot be disassembled, adj usted or replaced after the initial assembly . Due to the lack of configurability, independent single devices are less common and more application speci fic .

As to the particular scope of the present invention, there are no know previous developments to use optomechanical eye ( OME ) models for automotive display inspection with OME models . Known OME developments are only capable of analysing the characteristics of a "point spread function" through microscopy in a very small fraction of the image plane , in the microns order of magnitude to later compute the image quality of extended obj ects instead of evaluating them directly as is the obj ect of the present invention .

Therefore , the disclosed invention describes the elements of an optomechanical eye model ( OME ) , its routines and procedures for inspection of said obj ects , providing some preferred embodiments for implementation .

A typical eye model has an anterior lens system to represent the cornea, a circular hole or a diaphragm to simulate the pupil , a crystalline lens and/or an intraocular lens ( IOL ) , and a posterior viewing system which corresponds to the retina . In most devices , the relative position between some or all these components may be changed for fine tuning of the acquired images .

Current inspection routines for human user-oriented obj ects consider tight standards that might over-constrain the acceptance of pieces produced in industrial context . Whenever such obj ects will be subj ect to human eye inspection there is interest in being able to automate this process and use quality standards that are in line with the human eye detection capacity .

The possibility to automatically inspect obj ects , searching for defects that might be visible by the human eye under di f ferent physiological condition is one of the goals of the present invention . Also , it is considered the probability to evaluate how those will be seen under other human eye defects and anomalies . This will avoid the subj ectivity inherent to human observers , avoid tedious procedures for human operators , improve the reliability/consistency of the testing process and shorten the time cycle in quality control testing chains . Linked with automatic learning devices , this OME structure can be further improved by di f ferent machine learning fed by big data analysis in the context of industrial testing and progressively improve the ability to detect defects in said obj ects . The obj ect of the invention is therefore to provide a system with resolution common to subj ective inspection human eyes . In other words , the developed system returns a lower number of rej ection cases in test chain, which is compatible and equivalent with the rej ection rate of human inspector, compared to the higher rej ection rate of an obj ective system, but whose defects would not be seen under regular usability conditions .

Besides the subj ective inspection, obj ective arti ficial inspection devices might push the detection threshold well beyond what is seen by the human naked eye , resulting in higher rej ection cases in inspection chains . This might happen when the resolution of the camera is too high, optical components of the inspection system magni fy the image and/or the distance of inspection is closer than the usual usability distance , resulting in an arti ficially enhanced resolution that is not compatible with what is expected for the naked human eye .

Therefore , the herein disclosed invention describes a system configured to reproduce the human inspection chain conditions , that are more similar to the subj ective conditions of testing, resorting to the use of an arti ficial system that mimics the optical characteristics , the resolution of the detector and an adj ustable distance of inspection that is closer to the end user usability conditions .

A particular embodiment of the disclosed optomechanical eye system comprises aspheric surfaces to mimic cornea and whole eye aberrations in one single surface , and the possibility to include other devices such as intraocular lens (IOL) , retina with resolution equivalent to human eye. Automatic computer algorithms are used to control and change the optical conditions of the eye system, such as pupil size, axial length, lens geometry or optical aberrations, depending on the information acquired from the detected input. At physical level, for the proposed device it is possible to change the focus of crystalline lens with an optoelectronic lens and change the eye length by modifying the separation between the anterior and posterior segments. At an optical level, it is possible to control the shape of the cornea, the power of the internal lens and the artificial intraocular lenses.

Being the most important and critical refractive element of the human eye, and of extremely importance for the inspection procedures of the presently developed eye system, the complex optical design artificial cornea is fully customizable providing a surface that mimics the characteristics of the human eye under physiological or pathological conditions, providing an anterior corneal curvature between 4 and 11 mm of curvature (or any other value compatible with image formation) . The use of any corneal and crystalline lens surfaces asphericity / conic constants between +5.0 and -5.0 (or any other value compatible with image formation) is possible, or any standard or complex surface including, but not limited to bifocal devices, multifocal devices, asymmetric devices, refractive, diffractive or hybrid geometry, etc.

Beyond that, the corneal element is provided with exchangeable features, such as the shape, thickness or material of each lens surface, that allow to simulate different physiological and pathological conditions common in human observers/drivers . The internal structure of the eye system comprises a physiological refillable medium (water, water solution, saline medium, physiological solution, fluid gel or gelatine) , to simulate the internal conditions of the human eye and allowing the insertion of artificial prosthesis such as intraocular lenses. One possible embodiment is the insertion of a holder with an aperture (pupil) , where a replaceable IOL (ex: surgical IOL) may be attached. Different holders with different fixed or variable pupil sizes may also be used.

In one of the preferred embodiments of the present invention, the optomechanical eye (OME) system is provided with adjustable aperture, size and focal length of the lenses. The size of the OME is modified using a thread, but other methods to change it are also possible. A computer software may be added to precisely control the length of the device. The pupil size may also be computationally controlled when a variable aperture is used. The sensor resolution is able also to mimic the resolution capability of a human eye at the best-case scenario, i.e., within 60 cycles per degree. This might be achieved by adjusting different parameters of the optics and the resolution of the sensor including but not limited to: sensor resolution, eye length, design, geometry and power of the refractive surfaces of cornea and crystalline lens.

In another preferred embodiment of the present invention, the OME system is installed in a robotic arm to allow it to move in at least two perpendicular directions relatively to the target meant to be observed. Alternatively, the robotic arm(s) can move the target instead of moving the OME device, should the device remain standing still during the measurements in order not to disrupt the image acquisition . The referred robotic arm, or an additional robotic arm, is also able to rotate the OME device relatively to the center of the sensor . Both translation and rotation movements are meant to simulate head and eye movements .

The OME system is automatically or manually control led in a local or remote manner, through a software application commanded by a user or by other computational facilities including machine learning software . This can be achieved through controlling the robotic arm movement , and also the aperture , length and focus of the OME system, leading to a real time evaluation and detection of defects .

The detected defects are to be hierarchically categori zed as soon as possible , such classi fication determining the remaining of the analysis of the defect . Defects can be further subcategori zed and may change category i f enough evidence j usti fies so .

Feedback control is one of the proposed techniques that is used to adaptatively search for a defect or get more information about a found defect in order to help the classi fication process . This process is meant to mimic the psychophysical and physiological behaviour a human being would exhibit . For example , should a defect be present in the targeted image , the robotic arm may decrease the distance between the target and the sensor for a closer-up examination (head movement of a human observer ) . Di f ferent observation angles and accommodative conditions may also be considered (movements , gaze and accommodation of a human eye ) . Di f ferent algorithms will be used to analyse and categori ze di f ferent types of defects . The algorithm used depends on the information previously acquired by the defect . More generic algorithms are used at the beginning of the analysis and more speci fic ones may be used as soon as the defect is further characteri zed . Said algorithms may involve Hough trans forms , Hat trans forms , morphological filtering, high/band pass filtering, etc, or more elaborated methods such as neural networks . Targets to be analysed can also be compared with reference targets ( ex : by cross-correlation, matching filter, etc ) . Background light variations or direct illumination techniques can also be used to contaminate the output of the target display, in order to evaluate how the ability to spot defects is af fected when the display reflects light in addition to the light it generates , or when the OME/driver is disturbed by other sources of light .

Still with regard to prior art , the developed OME provides a solution for automotive display inspection that allows to increase the image analysis coverage from a few microns to an area that can go up to several millimetres or centimetres , depending on sensor si ze and resolution, this being achieved with a sensor that has a similar resolution to a human retina, contrary to other approaches that require much higher resolution devices or image magni fication through the use of optic settings . The suggested arrangement also does not require rotation to obtain of f-axis imaging information . These features can be achieved by customi zation of the parts involved which might comprise additive manufacturing ( 3-D printing) fabrication . Another key feature is related to present OME being able to work under both dry and wet condition by incorporating either aspheric sol id Planoconvex lenses or scleral lenses , respectively, all customi zable from the anterior and posterior surfaces to mimic the eye . To conclude , the customi zable corneal aspheric and other physiological or pathological complex surface profiles , combined with additive manufacture for easy prototyping and adaptation of components in a more realistic human-like way and an image analysis coverage area similar to the one proj ected in the human eye , allow to obtain a distinctive and innovative product when considering the technologies known till present date .

Brief description of the drawings

For better understanding of the present application, figures representing preferred embodiments are herein attached which, however, are not intended to limit the technique disclosed herein .

Figure 1 - depicts the mechanical details of the preferred embodiment of the optomechanical eye ( OME ) system in a transversal side view, wherein the reference numbers refer to :

100 - optomechanical eye system;

10 - anterior segment ;

15 - holder, comprising optical elements ;

20 - posterior segment ;

30 - camera casing, comprising a sensor .

Figure 2 - depicts the mechanical details of the preferred embodiment of the optomechanical eye ( OME ) system in a transversal side view, wherein the reference numbers refer to :

30 - camera casing, comprising a sensor ; 31 - optical CCD sensor die.

Figure 3 - depicts the mechanical details of the preferred embodiment of the optomechanical eye system (100) in a side view, wherein the reference numbers refer to:

100 - optomechanical eye system;

10 - anterior segment;

15 - Holder, comprising optical elements;

20 - posterior segment;

30 - camera casing, comprising a sensor.

Figure 4 - depicts the mechanical details of another preferred embodiment of the optomechanical eye system (100) in a side view with the extended adjusted thread, wherein the reference numbers refer to:

100 - optomechanical eye system;

10 - anterior segment;

15 - Holder, comprising optical elements;

20 - posterior segment;

30 - camera casing, comprising a sensor.

Figure 5 - depicts the mechanical details of a preferred embodiment of the posterior segment (20) in a transversal side view.

Figure 6 - depicts the mechanical details of the preferred embodiment of the posterior segment (20) in a side view.

Figure 7 - depicts the mechanical details of a preferred embodiment of the anterior segment (10) in a transversal side view where it is possible to glimpse a groove (14) . Figure 8 - depicts the mechanical details of the preferred embodiment of the anterior segment (10) in an upper transversal side view where it is possible to glimpse the groove ( 14 ) .

Figure 9 - depicts the mechanical details of the preferred embodiment of the anterior segment (10) in a side view where it is possible to glimpse the groove (14) .

Figure 10 - depicts the mechanical details of an alternative embodiment of the anterior segment (10) to simulate off-axis vision through an angular element (16) in an upper transversal side view where it is possible to glimpse the groove ( 14 ) .

Figure 11 - depicts the mechanical details of an alternative embodiment of the anterior segment (10) to simulate off-axis vision through an angular element (16) in an upper view where it is possible to glimpse the groove (14) .

Figure 12 - depicts the OME system assembled with the alternative embodiment of the anterior segment (10) to simulate off-axis vision through an angular element (16) .

Figure 13 - depicts the mechanical details of a preferred embodiment of the holder (15) in a transversal side view.

Figure 14 - depicts the mechanical details of the preferred embodiment of the holder (15) in a opposite transversal side view . Figure 15 - depicts the mechanical details of the preferred embodiment of the holder (15) enclosed within the anterior segment (10) in a transversal side view.

Figure 16 - illustrates the steps of the proposed method for object inspection (200) , wherein the reference numbers relate to:

201 - initialize;

202 - reset position (full size) ;

203 - acquire image (s) ;

204 - analyse data;

205 - found defect;

206 - (Re) Categorize defect;

207 - more detail;

208 - Final report;

209 - adjust the region of interest (ROI) ;

210 - move closer;

211 - keep searching;

212 - defect clear.

Description of Embodiments

With reference to the figures, some embodiments are now described in more detail, which are however not intended to limit the scope of the present application.

The herein disclosed invention describes, in one of the preferred embodiments, an optomechanical eye (OME) system (100) comprising at least 3 main components:

- A camera casing (30) , comprising a sensor located therein, which in a possible embodiment, is located at the bottom of said casing. The sensor may comprise one of a commercial camera attached to the device, or a more elaborated sensor, such as a curved sensor, optical fibers, etc;

- Anterior segment (10) ;

- Posterior segment (20) .

In one of the preferred embodiments of the present invention, the camera casing (30) comprises a digital acquisition camera, based on a C-mount camera, with a planar detecting region. Other mounting standards may be also used. The sensing area comprises a maximum length of 16 mm, and all components of the system comprise an internal diameter corresponding to said length. The sensor surface located inside the camera casing (30) is set at a predetermined distance from the posterior extremity of the posterior segment (20) accordingly to the C-mount standards, defining the flange focal distance.

The anterior segment (10) comprises a cylinder-shaped structure where optical elements (15) will be placed. The optical elements (15) , comprising a lens arrangement secured by a holder, referred as to cornea element, will be attached to the anterior segment (10) through a groove feature to insert the holder (15) with an additional Intraocular Lens (IOL) .

The posterior segment (20) ensures the connection between the anterior segment (10) , where the optic elements (15) are adapted, and the camera casing (30) through thread connection means, but other connection means may also be used. This posterior segment (20) also allows to change the axial length of the system (100) , by changing the overlapping length relatively to the anterior segment (10) , to simulate the effect of different refractive errors of the optomechanical eye system (100) . Said posterior segment (20) consists in a hollow cylinder which is inserted and adjusted to the camera casing (30) through screw means of the existing threads in both elements (20, 30) . The adjusting screwing distance leads to the anterior surface of the posterior segment (10) to be positioned at an exact known distance away from the sensor (30) surface, which turns it easier to measure and keep track of distances when adjusting the optics of the system (100) . If a different mount standard is used for the sensor, these dimensions may need to be adjusted in order to keep particular and exact distances, using a posterior segment (20) with a different length. Other mounting standards than C-mount can also be used, provided the posterior segment

(20) is adapted to it.

Internally, the posterior segment (20) has a predetermined diameter to match the camera casing (30) , and particularly the sensor sensing area. If a different sensing area is used, this diameter may need to be adjusted. A right-hand thread is placed in the anterior part of the posterior segment (20) to connect to the anterior segment (10) .

The anterior segment (10) is the front part of the system (100) . It comprises a predetermined length with an additional upper groove (14) , allowing to set maximum total length of the OME . An ISO metric profile thread allows the anterior segment (10) to be fitted into the posterior segment (20) , providing a minimum total length of the OME. Therefore, the total physical length of the system (100) can vary between set predetermined ranges, depending on the degree of entanglement between the anterior segment (10) and the posterior segment (20) . The outer radius is also set to match the same radius of the remaining components (this value may change if a different embodiment is used) . The anterior segment (10) comprises a groove (14) that transverses a portion of the outer ring-shaped segment (10) and allows to fit the holder comprised of the optical elements (15) . Other configurations may be used for different shapes of lens. The holder (15) comprises a transparent box with an intraocular lens inside. Such lens is attached to the aperture (pupil) of the holder. A small negative prominence on the posterior side of the holder allows the IOL to be attached by its haptics. The shape of the prominence may change, depending on the lens being attached. If the IOL is to be placed in front of the pupil instead, the holder can be flipped horizontally.

An alternative embodiment of the anterior segment (10) makes an angle at the optical centre (pupil) . Such configuration allows to rotate the optical axis around the optical centre to simulate off-axis vision. Different angles may be used.

It is to be stressed that other embodiments of the device are possible, where a different sensor and/or modified parts may be used instead of the ones mentioned above. As a consequence, the corresponding dimensions may be different.

The present developed optomechanical system (100) for automatic inspection of defects in objects, is also able to provide autonomous hierarchical classification and analysis of defects on displays. These defects are detected and evaluated in real time by the OME . Such classification will determine the remaining of the analysis of the defect. Defects can be further subcategorized and may change category if enough evidence justifies so.

Claims

The decisioning algorithms used to execute this task, are dependent of the previously acquired information of the defect. More generic algorithms are used at the beginning of the analysis and more specific and developed ones may be used as soon as an identified defect is further characterized. Algorithms used may involve Hough transforms, Hat transforms, morphological filtering, high/band pass filtering, etc, or more elaborated methods such as machine learning. Targets to be analysed can also be compared with reference targets (ex: by cross-correlation, matching filter, etc) . A feedback control software that controls the OME system movement and optical conditions, depending on the characteristics of what is being detected. A device and process of inspection that by adjusting the distance of observation and adjusting the region of interest to that of a potential defect, minimizes the possibility of false rejections while minimizing the possibility of false clearance of defective pieces. CLAIMS

1. Optomechanical system (100) for automatic inspection of defects in objects, comprising a camera casing (30) comprising an optical sensor located inside said casing (30) ; a posterior segment (20) of tubular hollow shape mechanically adjusted to said camera casing (30) through thread means comprised in an outer surface of the posterior segment (20) and comprised in an inner surface of an existing circular opening located in a side face of said casing (30) ; and an anterior segment (10) of tubular shape mechanically adjusted to said posterior segment (20) through thread means comprised in an outer surface the anterior segment (10) and comprised in an inner surface of the posterior segment (20) ; wherein the thread means of the mechanical adjustment between the anterior segment (10) and the posterior segment (20) comprise a predetermined length that allows to fine tune the distance between an anterior extremity of the anterior segment (10) and anterior extremity of the posterior segment (20) .

2. Optomechanical system (100) according to the previous claim, wherein the optical sensor is configured to be set at a predetermined distance from a posterior extremity of the posterior segment (20) accordingly to camera lens mounting standards, said distance defining the flange focal distance.

3. Optomechanical system (100) according to any of the previous claims, wherein the anterior segment (10) comprises a groove (14) in the anterior extremity to allow the mechanical fitting of a holder (15) which traverses a central portion of the tubular hollow shape.

4. Optomechanical system (100) according to any of the previous claims, wherein the anterior segment (10) comprises a tubular hollow shaped angular element (16) mechanically fitted between the anterior extremity and the thread means the anterior segment (10) .

5. Optomechanical system (100) according to any of the previous claims, wherein the holder (15) comprises an optoelectronic lens arrangement with fixed and/or variable apertures, adjusted through distancing of anterior and/or posterior segments separation.

6. Optomechanical system (100) according to any of the previous claims, wherein the fine tune of the distance between an anterior extremity of the anterior segment (10) and anterior extremity of the posterior segment (20) and the optoelectronic lens arrangement with variable apertures allows to capture and reproduce images of extended objects on a wide field of view within on-axis and off-axis and changing a distance of observation adjusting the region of interest (ROI) in case of a detection of non-obvious defects.

7. Method of operation (200) of the optomechanical system (100) for automatic inspection of defects in objects, according to any of the previous claims, comprising the steps of

System initialization (201) ;

Position reset of the acquired image to full size range (202) ;

Acquire image (203) ; Analyse acquired image data (204) ;

Categorizing defect (205) if a defect found (205) in the analysed acquired image data (204) , recategorizing said defect in case of recurrence through adjustment of the detail (207) and region of interest (209) of the acquired image (203) ;

Production of the final report (209) .

8. Method of operation (200) of the optomechanical system (100) for automatic inspection of defects in objects, according to the previous claim, wherein the categorizing defect (205) comprises additional searching (211) for defects in the acquired images (203) data (204) through adjustment of the position (210) of the optomechanical system (100) with regard to the object.