Defect detection system and method thereof for an optical film

Abstract

Disclosed herein is a system to detect one or more defects in an optical film comprising at least one illumination unit movable at one or more angles, at least one image acquisition unit movable at one or more angles, an optical film and a processor The processor configured to receive a first image of the optical film captured by the at least one image acquisition unit and classify the first image as related to one of a defective film or a defect-free film. The processor determines a confidence value of the classification exceeds a confidence threshold value and modifies one or more motion parameters of the optical film, the at least one illumination unit and / or the at least one image acquisition unit to capture one or more further images if the confidence value does not exceed the confidence threshold and detects one or more defects in the optical film by processing the one or more further images.

Description

[0001] DEFECT DETECTION SYSTEM AND METHOD THEREOF FOR AN OPTICAL FILM

[0002] FIELD OF THE DISCLOSURE

[0003] The present disclosure in general relates to optical films and more particularly, to a method and system for automated detection of defects of the optical film.

[0004] BACKGROUND

[0005] Optical thin films, particularly, optical prism films, are essential components in modern optical systems, used to manipulate light through interference and refraction. The optical thin films are laminated in a glass stack consisting of multiple other layers and hence also referred to as Light in Glass (LIG) stack or thin film stacks. These multiple layers may be composed of varying refractive indices, enabling precise control over light reflection, transmission, and absorption. One way of manufacturing the optical thin films may be using nano-imprinting of the thin film stacks. The performance of the optical thin films is critical in applications such as, but not limited to, automotive lighting and display systems. To ensure high-quality optical thin films, the optical thin films may be thoroughly inspected for any defects.

[0006] However, during nano imprinted thin film stack production, various uncertainties can lead to defects such as, but not limited to, scratches or marks, necessitating a time-consuming and resource-intensive quality control process. Furthermore, since the optical thin films are embedded inside the LIG stack, a quality engineer must inspect the entire stack, identify marks, and classify defects, extending the production process.

[0007] Thus, there is a need for a method and a system for an efficient and automated inspection of optical thin films that may fasten the defect detection process.

[0008] The information disclosed in this background of the disclosure section is only for enhancement of understanding of the general background of the disclosure and should not be taken as acknowledgment or any form of suggestion that this information forms prior art already known to a person skilled in the art.

[0009] SUMMARY

[0010] Disclosed herein is a system to detect one or more defects in an optical film comprising at least one illumination unit movable at one or more angles, at least one image acquisition unit movable at one or more angles and an optical film mounted at an angle relative to the at least one illumination unit and the at least one image acquisition unit. The system further comprises a processor communicatively coupled with the optical film, the at least one illumination unit and the at least one image acquisition unit. The processor is configured to receive a first image of the optical film captured by the at least one image acquisition unit and classify the first image as related to one of a defective film or a defect-free film. The processor is configured to determine a confidence value of the classification exceeds a confidence threshold value to detect one or more defects related to the defective film upon classifying the first image as related to the defective film. The processor is configured to modify one or more motion parameters of one or more of the optical film, the at least one illumination unit and the at least one image acquisition unit to capture one or more further images of the optical film upon determining that the confidence value does not exceed the confidence threshold. The processor is configured to receive the one or more further images captured by the at least one image acquisition unit based on the modified one or more motion parameters. The processor is configured to recursively determine the confidence value of classification of the one or more further images, modify the one or more motion parameters upon determining the confidence value does not exceed the confidence threshold and receive the one or more further images until the confidence value exceeds the confidence threshold. Further, the processor is configured to detect one or more defects in the optical film by processing the one or more further images.

[0011] Also disclosed herein is a system to detect one or more defects in an optical film comprising two image acquisition units configured to capture one or more images of the optical film. A first image acquisition unit mounted on a first side of the optical film and configured to capture one or more images of the optical film from the first side of the optical film. A second image acquisition unit mounted at a second side of the optical film and configured to capture one or more images of the optical film from the second side of the optical film. The second side is an opposite side of the optical film. The system comprises two illumination units configured to illuminate the optical film. A first illumination unit mounted at an acute angle from an axial plane of the optical film and a second illumination unit mounted perpendicular to an axis of the axial plane of the optical film. The system further comprises an image processing unit coupled with the two image acquisition units and configured to process the one or more images to detect one or more defects in the optical film. The system also comprises one or more positioning units coupled with one or more of the two image acquisition units, the two illumination units and the optical film and configured to control a movement of one or more of the optical film, the two image acquisition units, the two illumination units. The system comprises a processor coupled with the two image acquisition units, the two illumination units, the one or more positioning units and the image processing unit. The processor is configured to control at least a movement of one or more of the two image acquisition units, the two illumination units, the optical film to capture one or more images of the optical film. The processor is configured to detect one or more defects using a machine learning model by processing the one or more images and mapping the one or more defects on the one or more images. The processor is configured to transmit one or more notifications related to the one or more defects to a user device.

[0012] Further disclosed herein is a method for detecting one or more defects in an optical film. The method comprises receiving a first image of the optical film captured by the at least one image acquisition unit and classifying the first image as related to one of a defective film or a defect- free film. The method comprises determining a confidence value of the classification exceeds a confidence threshold to detect one or more defects related to the defective film upon classifying the first image as related to the defective film. The method comprises modifying one or more motion parameters of one or more of the optical film, at least one illumination unit and the at least one image acquisition unit to capture one or more further images of the optical film upon determining that the confidence value does not exceed the confidence threshold. The method further comprises receiving the one or more further images captured by the at least one image acquisition unit based on the modified one or more motion parameters. Further, the method comprises recursively determining the confidence value of classification of the one or more further images, modifying the one or more motion parameters upon determining the confidence value does not exceed the confidence threshold and receiving the one or more further images until the confidence value exceeds the confidence threshold. The method also comprises detecting one or more defects in the optical film by processing the one or more further images.

[0013] The foregoing summary is illustrative only and is not intended to be in any way limiting. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features will become apparent by reference to the drawings and the following detailed description. BRIEF DESCRIPTION OF DRAWINGS

[0014] The accompanying drawings, which are incorporated in and constitute a part of this disclosure, illustrate exemplary embodiments and, together with the description, serve to explain the disclosed principles. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. The same numbers are used throughout the figures to reference like features and components. Some embodiments of device or system and / or methods in accordance with embodiments of the present subject matter are now described, by way of example only, and with reference to the accompanying figures, in which:

[0015] Fig- 1 illustrates an exemplary architecture of a system to detect one or more defects in an optical film;

[0016] Fig- 2 illustrates an exemplary schematic diagram of the optical film in accordance with embodiments of the present disclosure;

[0017] Fig- 3 illustrates an exemplary architecture of the system to detect one or more defects of the optical film in accordance with an embodiment of the present disclosure;

[0018] Fig. 4 illustrates an exemplary architecture of the system to detect one or more defects of the optical film in accordance with another embodiment of the present disclosure;

[0019] Fig. 5 illustrates an exemplary architecture of the system to detect one or more defects of the optical film in accordance with yet another embodiment of the present disclosure;

[0020] Fig. 6 illustrates an exemplary flowchart of a method for detecting one or more defects of the optical film in accordance with an embodiment of the present disclosure; and

[0021] Fig. 7 illustrates an exemplary flowchart of a method for detecting one or more defects of the optical film in accordance with another embodiment of the present disclosure.

[0022] The figures depict embodiments of the disclosure for purposes of illustration only. One skilled in the art will readily recognize from the following description that alternative embodiments of the structures and methods illustrated herein may be employed without departing from the principles of the disclosure described herein.

[0023] DETAILED DESCRIPTION In the present document, the word "exemplary" is used herein to mean "serving as an example, instance, or illustration." Any embodiment or implementation of the present subject matter described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments.

[0024] While the disclosure is susceptible to various modifications and alternative forms, specific embodiment thereof has been shown by way of example in the drawings and will be described in detail below. It should be understood, however that it is not intended to limit the disclosure to the particular forms disclosed, but on the contrary, the disclosure is to cover all modifications, equivalents, and alternative falling within the scope of the disclosure.

[0025] The terms “comprises”, “comprising”, or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a setup, device or method that comprises a list of components or steps does not include only those components or steps but may include other components or steps not expressly listed or inherent to such setup or device or method. In other words, one or more elements in a device or system or apparatus proceeded by “comprises. . . a” does not, without more constraints, preclude the existence of other elements or additional elements in the device or system or apparatus.

[0026] The present disclosure relates to various methods and systems to detect one or more defects in an optical film. The system comprises one or two illumination units, one or two image acquisition units and a processor. The illumination units are movable and are configured to illuminate the optical film in one or more angles. The image acquisition units are movable and configured to capture one or more images of the optical film at one or more angles. The processor is configured to capture one or more images of the optical film and classify the optical film as a defective film or a defect-free film. If the optical film is classified as the defective film, the system may detect if the one or more captured images are sufficient to detect one or more defects within the defective film. If the one or more captured images are not sufficient, the system may recursively adjust the positions of the image acquisition units and / or the illumination units to capture further images based on types of defects of locations of defects.

[0027] Further, the system may verify whether the one or more further images are sufficient to detect one or more defects within the defective film. If the system determines that the one or more further images are sufficient, the system may compare the one or more further images with a plurality of defective film images within a defect database and may detect the one or more defects. Further, and notify a user indicating the positions and other parameters of the one or more defects on the optical film. Thus, the system detects the defects within the optical film using only one or two illumination units and one or two image acquisition units, thereby reducing the hardware requirements. Further, the system also illuminates the optical film and / or acquires images at optimum angles, thereby optimizing the inspection process.

[0028] In the following detailed description of the embodiments of the disclosure, reference is made to the accompanying drawings that form a part hereof, and in which are shown by way of illustration specific embodiments in which the disclosure may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the disclosure, and it is to be understood that other embodiments may be utilized and that changes may be made without departing from the scope of the present disclosure. The following description is, therefore, not to be taken in a limiting sense.

[0029] Fig- 1 illustrates an exemplary architecture of a system to detect one or more defects in an optical film.

[0030] As shown in Fig. 1, the exemplary architecture 100 comprises a defect detection system (DDS) 102 to detect one or more defects of an optical film 104. The architecture 100 may also comprise at least a defect database 106 and a user device 108 of a user 109. The DDS 102 may be communicatively coupled with the defect database 106 and the user device 108 using a wired or a wireless network connection (indicated as a solid line), such as, but not limited to, Local Area Network or the Internet. The DDS 102 may communicate with the optical film 104 in a manner such as, but not limited to, illuminating the optical film 104 and / or capturing one or more images of the optical film 104. The DDS 102 may comprise one or more components to detect one or more defects of the optical film 104. The one or more components may comprise at least one illumination unit 110, at least one image acquisition unit 112, a processor 114 and a communication unit 116. The processor 114 is communicatively coupled with the at least one illumination unit 110, the at least one image acquisition unit 112, and the communication unit 116. In some embodiments, the processor 114 may be configured at a centralized server away from the DDS 102 and may be communicatively coupled with the DDS 102 through a communication network such as, but not limited to, the Internet.

[0031] The optical film 104 may be a thin film comprising a stack of a plurality of layers to carry light within the optical film 104. The plurality of layers may comprise at least a transparent layer, a reflective layer, a metallic coating layer, and an adhesive layer. The transparent layer may be a base layer deployed to provide structural support to the optical film 104 and may be made of an optically transparent material that may transmit light through the layer. The transparent layer may be made of materials such as, but not limited to, glass or a type of plastic. In one example, the transparent layer may be composed of Polyethylene terephthalate (PET). The reflective layer may be reflect light within the layer and may comprise of a plurality of nano imprinted reflective structures such as, but not limited to, a plurality of prism-shaped nano reflective structures.

[0032] The metallic coating layer may be a coating made of a metallic material and may be coated on top of the reflective layer to guide the light within the reflective layer. The metallic coating layer may reduce reflections and may increase an amount of light transmitted through the nano reflective structures. In some embodiments, the metallic coating layer may be made of metallic material such as, but not limited to, Gold, Copper, Silver, Aluminium and the like. The adhesive layer may be located above the metallic coating layer and may be made of an optically clear material. The layers of the optical film 104 described herein are only for illustrative purposes and cannot be construed as limiting in any manner. The optical film 104 may comprise more layers or less layers compared to the layers described herein. In an embodiment, the optical film 104 may be of approximately 100-150 microns of thickness.

[0033] In some embodiments, the optical film 104 may be composed of a long sheet of film rolled using a rolling unit. The rolling unit may comprise at least a winding roller, an unwinding roller and a motor. The motor may be configured to control the movement of the winding and the unwinding rollers. The unwinding roller may unwind the optical film 104 for detection of one or more defects and the winding roller may wind up the optical film 104 upon the detection. Each of the winding roller and the unwinding roller may be made of one or more chemically inert materials in accordance with some embodiments of the present disclosure. The one or more chemically inert materials may comprise, but not limited to, Silicon and rubber. Using the chemically inert materials for the winding and the unwinding rollers may prevent contamination or damage to the optical film 104 during defect detection, alternatively referred to herein as ‘inspection’.

[0034] Further, the rollers made of the chemically inert materials may facilitate a smooth movement of the optical film 104 between the rollers and may avoid any defects caused due to rough surfaces of the rollers. In these embodiments, the movement of the winding roller and the unwinding roller may be achieved using at least a motor coupled with the winding roller and the unwinding roller. In some embodiments, the winding and the unwinding rollers may be made of one or more self-healing materials comprising a multi-layer polymer composition. The multi-layer polymer composition may comprise at least an outer protective layer and a core structure layer. The outer protective layer may be composed of chemically inert material with self-healing properties to heal one or more defects occurring due to mechanical stress or surface abrasion. The core structural layer may be composed of a metallic material to provide mechanical stability ensuring uniform performance under heavy-load scenarios.

[0035] In other embodiments, the optical film 104 may be a small optical film that may be placed on a movable table configured to move the small optical films during the inspection. In some other embodiments, the optical film 104 may be rolled using the unwinding roller on one side and a slicing device on the other side that may be configured to cut the optical film 104 into small films upon inspection. For example, the slicing device may cut and discard a portion of the optical film 104 that may comprise of the one or more defects. Further, the optical film 104 may be rolled or moved horizontally, vertically, or at any other angle during the inspection.

[0036] The defect database 106 may comprise a plurality of datasets required to train one or more machine learning (ML) models or for processing by one or more image processing modules of the DDS 102. The plurality of datasets may comprise a preliminary list of defects, impact of each defect, a plurality of defective film images and a plurality of defect-free images. The plurality of defective film images may comprise images corresponding to a plurality of defects a plurality of defective optical films. The plurality of defect-free images may comprise images corresponding to a plurality of defect-free optical films. In one embodiment, the plurality of datasets may also comprise various classes of defects related to, without limiting to, the transparent layer, the reflective layer, the metallic coating layer, the adhesive layer, the film, functional defects and environmental defects. Table 1 illustrates an exemplary preliminary list of defects comprising classes, sub-classes, causes and impacts of defects on the performance of the optical film 104.

[0037]

[0038] Table 1

[0039] The preliminary list of defects illustrated in Table 1 has only been described for illustrative purposes and cannot be construed as limiting in any manner. Further, the defect database 106 may receive inputs from the user 109 or from the DDS 102 to update the preliminary list of defects and / or to store one or more images of new defects or one or more new images of a defect of the preliminary list and accordingly update the defect database 106. In some embodiments, the DDS 102 may update the defect database 106.

[0040] The user device 108 may be a computing device of the user 109 that is configured to receive one or more notifications from the DDS 102 related to the one or more defects and display the one or more notifications to the user 109. The user device 108 may be any computing device such as, but not limited to, a smart phone, a smart watch, a tablet device, a laptop, a palmtop, or a desktop device. The user 109 may include, but not limited to, an inspection professional who may inspect the one or more defects of the optical film 104 and may take actions on the optical film 104 in response to the one or more defects.

[0041] The user device 108 may be configured with a transceiver to receive the one or more notifications from the DDS 102 and a display device to display the one or more notifications to the user 109. The user device 108 may also comprise a plurality of other components that have not been described herein. In one embodiment, the user device 108 may comprise an input device to receive one or more inputs from the user 109 regarding the one or more defects and may transmit the one or more inputs to a manufacturing plant of the optical film 104. In one example, the one or more inputs may be one or more processing parameters of the manufacturing plant required to be modified to avoid the occurrence of the one or more defects in a further optical film 104. In other embodiments, the user 109 may be a manufacturing professional receiving the one or more defects of the optical film 104 and the one or more processing parameters from the DDS 102.

[0042] The at least one illumination unit 110 may comprise one or more light sources configured to illuminate the optical film 104. Each light source may be configured to generate light of various wavelengths such as, but not limited to, visible light, Laser light, Infrared (IR) light, Ultraviolet (UV) light and the like. The at least one illumination unit 110 may be movable around the optical film 104 to illuminate the optical film 104 in desired angles. In one embodiment, the at least one illumination unit 110 may be coupled with at least one positioning unit configured to move the at least one illumination unit 110.

[0043] In some embodiments, the at least one illumination unit 110 may comprise only one light source and may be rotated and / or revolved around the optical film 104 to illuminate the optical film 104 at any angle and at any position. In these embodiments, the light source may be movable from 0°-360° from the optical film 104 in a spherical coordinate system. In some other embodiments, the at least one illumination unit 110 may comprise two illumination units such as a first illumination unit and a second illumination unit. The first illumination unit may be mounted at an acute angle from an axial plane of the optical film 104 and the second illumination unit may be mounted perpendicular to an axis of the axial plane of the optical film 104. In one example, the first illumination unit may be mounted at an angle of 45° from the axial plane or horizontal plane of the optical film 104 and the second illumination unit may be mounted at an angle perpendicular to an X-axis of the optical film 104.

[0044] The at least one image acquisition unit 112 may comprise one or more image sensors configured to capture one or more images of the optical film 104. The image sensor may comprise various types of sensors, but not limited to, visible light based sensor, IR sensor, UV sensor, Laser based image sensor and the like. In an embodiment, the at least one image acquisition unit 112 may be a line scan camera. The at least one image acquisition unit 112 may be movable around the optical film 104 to capture the one or more images of the optical film 104 in desired angles. In one embodiment, the at least one image acquisition unit 112 may be coupled with at least one positioning unit configured to move the at least one image acquisition unit 112.

[0045] In some embodiments, the at least one image acquisition unit 112 may comprise only one image sensor and may be rotated and / or revolved around the optical film 104 to capture the one or more images the optical film 104 at any angle and at any position. In these embodiments, the image sensor may be movable from 0°-360° from the optical film 104 in a spherical coordinate system. In some other embodiments, the at least one image acquisition unit 112 may comprise two image acquisition units such as a first image acquisition unit and a second image acquisition unit. The first image acquisition unit may be configured to capture the one or more images of a first side of the optical film 104 and the second image acquisition unit may be configured to illuminate a second side of the optical film 104. The second side may be an opposite side of the first side. In an example, the first and second sides may be upper and lower portions of the optical film 104 when the optical film 104 is inspected in a horizontal position. In these embodiments, each of the first and the second image acquisition units may be configured to rotate 0°-180° on the first and the second sides respectively of the optical film 104.

[0046] The processor 114 may be configured to process the one or more images captured by the at least one image acquisition unit 112 and may detect one or more defects of the optical film 104. The processor 114 may be configured with a defect detection model 118 trained to detect the one or more defects of the optical film 104. The defect detection model 118 may be a machine learning (ML) model trained using the plurality of datasets of the defect database 106, which is explained hereafter in detail. Further, the processor 114 may generate one or more notifications of the one or more defects and may transmit the one or more notifications to the user device 108 through the communication unit 116.

[0047] The communication unit 116 may be configured to communicate with the user device 108. The communication unit 116 may comprise at least a transceiver configured to transmit information to and / or receive information from the user device 108. The communication unit 116 may be coupled with the processor 114 to transmit one or more notifications to the user device 108 and receive one or more user inputs from the user device 108 and provide the one or more user inputs to the processor 114.

[0048] In operation, the processor 114 may be configured to curate and update the defect database 106. To curate the defect database 106, the processor 114 may capture a plurality of images of a plurality of optical films and may annotate each of the plurality of images as related to one of a defective image and a defect-free image based on a user input. For example, the user input may be received from the user 109. During annotation as the defective image, the processor 114 may annotate a class of defect of the defective image based on the user input. The various classes of the defects may be related to, without limiting to, the transparent layer, the reflective layer, the metallic coating layer, the adhesive layer, the optical film, the functional defects and the environmental defects.

[0049] Upon annotating based on the class, the processor 114 may also annotate a sub-class of the class of the defect based on the user input. For example, the sub-class may be a pinhole for a class of the metallic coating layer. Upon annotation, the processor 114 may store the plurality of defective film images, the plurality of defect-free images, classes, sub-classes, potential causes and impact of defects on performance on the optical film 104. The processor 114 may further update the defect database 106 when a new defect is detected in the optical film 104 that is not stored in the defect database 106 or when a new image of an existing defect is detected in the optical film 104.

[0050] Upon curating the defect database 106, the processor 114 may train the defect detection model 118 with the plurality of datasets of the defect database 106. The processor 114 may provide the plurality of datasets as input to the defect detection model 118, also referred to herein as a model 118 for training including testing and validation. In some embodiments, the model 118 may be one of a linear regression model, a decision tree model or a neural network model and the like. The model 118 may analyze each image of the plurality of datasets and may predict a classification for each image as related to the defective image or the defect-free image. In case of the defective image, the model 118 may further classify each image based on the class and the sub-class. Further, the processor 114 may analyze the predictions and may modify the model 118 based on the predictions to train the model 118. The processor 114 may also evaluate an accuracy of each prediction, compare the accuracy with a desired accuracy and may continue to train the model 118 until the accuracy is same or exceeds the desired accuracy. The desired accuracy may be defined as an accuracy of the prediction provided by a user, for example, the user 109, as input. Thus, the model 118 may learn one or more image features of the plurality of defective images and the plurality of defect-free images.

[0051] In some embodiments, the model 118 may be trained to classify each image as each of the defective film, classes and sub-classes of the defective film and defect-free film with respective confidence values. For example, the model 118 may classify an image as defective image with 70% confidence value whereas a defect-free image with a confidence value of 10%. The model 118 may also classify that the defective image is related to the transparent layer with 60% confidence, the reflective layer with 20% confidence and other types of classes with 5% confidence each. Based on this prediction, the processor 114 may determine that the image is a defective image with a defect in the transparent layer.

[0052] The processor 114 may initiate inspection of the optical film 104 using one or more images received from the at least one image acquisition unit 112. Initially, the at least one illumination unit 110 may illuminate the optical film 104 at one or more of the acute angle and the angle perpendicular to an axis of the axial plane of the optical film 104. Upon illuminating the optical film 104, the at least one image acquisition unit 112 may capture a first image of the optical film 104 at an initial angle. The initial angle may include, but not limited to, 90° from the axial plane or 45° from the axial plane. In some embodiments, the at least one image acquisition unit 112 may capture a plurality of first images at one or more initial angles. The model 118 may classify each of the plurality of first images as related to the defective film or a defect-free film.

[0053] Further, the processor 114 may provide the first image to the model 118 to predict a classification of optical film 104 and the confidence value of the classification. The model 118 may evaluate a similarity between the first image and a plurality of defective film images, and a plurality of defect-free images. The processor 114 may determine a similarity between the first image and at least one of the plurality of defect-free film images with a first threshold value. For example, the first threshold value may be 90%. If the similarity exceeds the first threshold value, the processor 114 may classify the first image as related to the defect-free film. Alternatively, the processor 114 may determine a similarity between the first image and at least one of the plurality of defective film images with a second threshold value. For example, the first threshold value may be 80%. If the similarity exceeds the second threshold value, the processor 114 may classify the first image as related to the defective film. In some embodiments, the first and second threshold values may be same.

[0054] In other embodiments, the model 118 may detect a similarity between the first image and each of the plurality of defective film images and may compare the similarity with a similarity threshold value. If the similarity does not exceed the similarity threshold value, the model 118 may classify the first image as defect-free image. If the similarity exceeds or same as the similarity threshold value with at least one of the plurality of defective film images, the model 118 may classify the first image as the defective film image.

[0055] In some embodiments, the processor 114 may be configured to perform one or more image processing operations to classify the first image as related to defective film or a defect-free film instead of the model 118. In these embodiments, the processor 114 may perform one or more image processing operations, evaluate the similarity between the first image and each of the plurality of defective images and the plurality of defect-free images and may classify the first image as related to the defective image or the defect-free image based on the similarity.

[0056] In an embodiment, the classification of the an input image as related to a defective film or a defect-free film may comprise the processor 114 pre-processing the input image to remove noise from the input image. The pre-processing may comprise one or more image processing operations including, but not limited to, gray-scale conversion, Gaussian blur, Canny edge detection, and contour masking. Upon removing the noise, the processor 114 may identify one or more Regions of Interest (ROIs) from the pre-processed input image. The processor 114 may provide the one or more ROIs to the model 118 to detect the one or more defects. The model 118 may compare the each of the one or more ROIs with each of a plurality of defective film images. The plurality of defective film images may be defined as a plurality of images that correspond to a plurality of sub-classes of defects of a plurality of classes of defects of the input image. Further, the model 118 may evaluate a similarity between each of the one or more ROIs with each of the plurality of defective film images. For example, the similarity of an ROI may be 80% with a defective film image among the plurality of defective film images. Further, the model 118 may compare the similarity with a similarity threshold value. The similarity threshold value may be defined as a value indicating a required level of similarity between the ROI and the defective film image to classify the defect. For example, the similarity threshold value may be 60%. The similarity threshold value may be provided by the user 109 or may be set by the model 118 or the processor 114 automatically.

[0057] If the similarity of an ROI compared to at least one defective film image of a sub-class exceeds the similarity threshold value, the model 118 may classify the ROI as the sub-class of the defect. For example, if the ROI matches with a defective film image of a ‘delamination’ sub-class of the ‘optical film’ class, the model 118 may classify the ROI as the ‘delamination’ defect.

[0058] If the model 118 classifies the first image as a defect-free image or related to a defect-free film, the processor 114 may perform a plurality of predefined inspection patterns to verify if the model 118 still classifies the optical film 104 as defect-free film. The plurality of predefined inspection patterns may include, but not limited to, moving positions and / or angles of one or more of the at least one illumination unit 110, the at least one image acquisition unit 112, the optical film 104 in a pre-defined manner. In one embodiment, the plurality of predefined inspection patterns may comprise a plurality of steps including, but not limited to, changing one or more illumination parameters of the at least one illumination unit 110. The one or more illumination parameters may include, but not limited to, a level, an intensity, a color and an incident angle of light.

[0059] The plurality of steps may include changing one or more motion parameters of the at least one image acquisition unit 112 comprising the angle, a type of the image sensor. In other embodiments, the plurality of predefined inspection patterns may include any other inspection processes that may be known to a person skilled in the art. Each of the captured plurality of images may be further classified by the model 118 as one of the defective image or the defect- free image. At the end of the plurality of predefined inspection patterns, if the model 118 still classifies the optical film 104 as a defect-free film, the optical film 104 may be considered as successfully tested.

[0060] Alternatively, if the model 118 classifies the first image as a defective film image or related to a defective film, the processor 114 may compare the confidence value with a confidence threshold. The confidence value may be evaluated based on the similarity. For example, the higher the similarity, the greater may be the confidence value. The confidence threshold may be any value required to detect whether the first image is sufficient to determine one or more defects of the optical film 104. For example, the confidence threshold may be 60%. If the confidence value exceeds the confidence threshold, the processor 114 may detect one or more defects in the optical film 104 using the first image.

[0061] Alternatively, if the confidence value of the class of defect does not exceed the confidence threshold, the processor 114 may modify one or more motion parameters of one or more of the at least one illumination unit 110, the at least one image acquisition unit 112, the optical film 104 based on the class. Upon modifying the one or more motion parameters, the processor 114 may acquire one or more further images to classify the defect with a greater confidence value. In other words, the processor 114 may modify a position or configuration of one or more of the at least one illumination unit 110, the at least one image acquisition unit 112, the optical film 104 to capture the one or more further images. For example, if the class of defect is the transparent layer, the processor 114 may position the at least one illumination unit 110 to highly illuminate the transparent layer. In one embodiment, the processor 114 may gradually move the at least one illumination unit 110 and the at least one image acquisition unit 112 at predefined sets of degrees, such as, but not limited to multiples of 5 or 10, to capture the one or more further images. Further, the processor 114 may position the at least one image acquisition unit 112 at angle such that the at least one image acquisition unit 112 may capture one or more further images of the transparent layer.

[0062] The one or more motion parameters of the at least one illumination unit 110 may include, but not limited to, an intensity, a wavelength, a position and an angle of the at least one illumination unit 110. The one or more motion parameters of the at least one image acquisition unit 112 may include, but not limited to, a position, the type of image sensor and an angle of the at least one image acquisition unit 112 to capture images.

[0063] The one or more motion parameters of the optical film 104 may include, but not limited to, a position and an angle of the optical film 104, a movement of a table moving the optical film 104. In an embodiment, where the optical film 104 may be rolled by the winding and the unwinding rollers, the processor 114 may modify one or more motion parameters of a motor that controls the motion of the rollers. The one or more motion parameters of the motor may include, but not limited to, a speed of the motor. For example, the processor 114 may reduce the speed of the motor to capture the one or more further images of the optical film 104.

[0064] Upon capturing the one or more further images, the model 118 may evaluate a similarity between each of one or more ROIs of the each further image with each of a set of defective film images. The set of defective film images may be defined as a plurality of images that correspond to a plurality of sub-classes of defects associated with the class of defect of the first image. For example, if the class of defect is of functional defects, the model 118 may compare the one or more ROIs with the set of defective film images annotated as various sub-classes of the functional defects. In an example, the similarity of an ROI may be 80% with a defective film image among the set of defective film images. Further, the model 118 may compare the similarity with the similarity threshold value.

[0065] If the similarity of an ROI compared to at least one defective film image of a sub-class exceeds the similarity threshold value, the model 118 may classify the ROI as the sub-class of the defect. On the other hand, if the similarity of an ROI compared to any of the set of defective film images does not exceed the similarity threshold value, the model 118 may classify the ROI as a new defect. Similarly, the model 118 may classify each of the one or more ROIs of each of the one or more further images as one of ‘sub-class of defect corresponding to the class of defect’ or a ‘new defect’.

[0066] Further, the processor 114 may map the one or more defects of the one or more ROIs on the further image and may transmit one or more notifications related to the one or more defects to the user device 108. The mapping may include a process of indicating locations, also referred to herein as ‘indications’ of the one or more defects on the further image to indicate the locations of the one or more defects to the user 109. Each notification may include, but not limited to, the mapped image, the class, the sub-class, the potential cause and the impact of the defect on the performance of the optical film 104. The processor 114 may transmit the one or more notifications to the user device 108. The user 109 may analyze the one or more notifications and may determine whether each of the one or more defects are minor or major. If the defect is a minor defect, the optical film 104 may be provided to further modules for processing of the optical film 104. The further modules may include, but not limited to, packing of the optical film 104 to a customer. If the defect is a major defect, the optical film 104 may be discarded and may not be provided to the further modules.

[0067] Further, the user 109 may determine one or more process parameters that may be modified during the manufacturing of further optical films and to avoid the one or more defects in the further optical films. In some embodiments, the processor 114 may determine the one or more process parameters and may transmit the one or more process parameters within the one or more notifications to the user device 108.

[0068] If the defect is a ‘new defect’, the notification may include the mapped image and may prompt the user 109 to provide the inputs to classify the ‘new defect’ . Further, the user 109 may analyze the one or more notifications and may include more details about the new defect such as, but not limited to, class, sub-class, potential causes and impact of the defect on the performance of the optical film 104. The processor 114 may receive the inputs and may update the defect database 106 with the new defect and the details of the new defect provided by the user 109.

[0069] Thus, the DDS 102 may provide real time defect detection ensuring high quality prism films may be utilized for further processing so that the product integrity can be verified. Further, the automated inspections of the DDS 102 may significantly reduce manual inspection times for larger rolls of optical films 104. Inspection of nanoimprinted prisms implies the system is useful for complex stacks with high-resolution capabilities which will have significant reduction in manual inspection times for large rolls of prism films. The DDS 102 may enable manufacture of high quality optical films that may ensure improved dissemination of light towards visibility of patterns within the LIG stack. The DDS 102 may also improve process of manufacturing the optical films 104, quality control and maintaining the product integrity by early detection of defects in the optical film 104 even before lamination in LIG stack.

[0070] Further, the DDS 102 may also be integrated in manufacturing production lines of the optical films to ensure high-quality defect free optical films. The DDS 102 may be customized and for any type of optical films with stacked layers including, including, but not limited to, the stack optical films with different coatings and layered glass stacks. Further, the curation and regular updation of the defect database 106 may also enable accurate and effective categorization of defects.

[0071] Fig- 2 illustrates an exemplary schematic diagram of the optical film 104 in accordance with embodiments of the present disclosure.

[0072] As shown in Fig. 2, the schematic diagram 200 of the optical film 104 illustrates a plurality of layers 202, 204, 206 and 208. The layer 202 may be the transparent layer, the layer 204 may be the reflective layer, the layer 204 may be the metallic coating layer and the layer 208 may be the adhesive layer. The reflective layer 204 may comprise a plurality of prism shaper nanostructures 210. Fig- 3 illustrates an exemplary architecture of the DDS 102 to detect one or more defects of the optical film 104 in accordance with an embodiment of the present disclosure.

[0073] As shown in Fig. 3, the architecture 300 comprises two illumination units 110-1 and 110-2 and two image acquisition units 112-1 and 112-2. The illumination unit 110-1 may be mounted at an acute angle, for example 45° from the axial plane of the optical film 104. The illumination unit 110-2 may be mounted at an angle perpendicular to an X-axis of the optical film 104. The image acquisition unit 112-1 may be mounted above the optical film 104 and the image acquisition unit 112-2 may be mounted below the optical film 104. Each of the two illumination units 110-1 and the two image acquisition units 112 may be coupled with a positioning unit (not shown in Fig. 3) that may be configured to move the units. In one example, the positioning units may be actuators. The two image acquisition units 112-1 and 112-2 may be movable between 0°-180° angles. The illumination unit 110-1 may be movable between 0°-180° angles above the optical film 104 and the illumination unit 110-2 may be movable along an edge of the optical film 104 or between 0°-180° angles along the edge.

[0074] Fig- 4 illustrates an exemplary architecture of the DDS 102 to detect one or more defects of the optical film 104 in accordance with another embodiment of the present disclosure.

[0075] As shown in Fig. 4, the DDS 102 may comprise only one image acquisition unit 112 that may be movable between 0°-360° angles and may capture the one or more images or one or more further images by rotating the image acquisition unit 112 at any angle and any position.

[0076] Fig- 5 illustrates an exemplary architecture of the DDS 102 to detect one or more defects of the optical film 104 in accordance with yet another embodiment of the present disclosure.

[0077] As shown in Fig. 5, the optical film 104 may be a long sheet of film rolled between an unwinding roller 502 and a winding roller 504 from right side to a left side. The winding roller 504 and the unwinding roller 502 may be coupled with a motion system 506. The motion system 506 may comprise at least a motor and a motion sensor. The motor may be configured to facilitate motion to the winding and the unwinding rollers. The motion sensor may be configured to sense a motion speed of the rollers 502 and 504. The DDS 102 may illuminate the optical film 104 using two illumination units 110-1 and 112-2 and two image acquisition units 112-1 and 114-2. The processor 114 may be configured to modify the one or more motion parameters of the motor and may control the motion of the optical film 104.

[0078] Fig. 6 illustrates an exemplary flowchart of a method for detecting one or more defects of the optical film 104 in accordance with an embodiment of the present disclosure.

[0079] The method 600 may be described in the general context of computer executable instructions. Generally, computer executable instructions can include routines, programs, objects, components, data structures, procedures, modules, and functions, which perform specific functions or implement specific abstract data types. The order in which the method 400 is described is not intended to be construed as a limitation, and any number of the described method blocks can be combined in any order to implement the method. Additionally, individual blocks may be deleted from the methods without departing from the scope of the subject matter described herein. Furthermore, the method can be implemented in any suitable hardware, software, firmware, or combination thereof.

[0080] At block 602, the DDS 102 may receive a first image of the optical film 104 captured by the at least one image acquisition unit 112.

[0081] At block 604, the DDS 102 may classify if the first image is related to a defective film or a defect-free film. If the first image is related to a defective film, the DDS 102 may proceed to block 606 and if the first image is related to a defect-free film, the DDS 102 may proceed to block 608.

[0082] At block 606, the DDS 102 may determine if the confidence value of the classification exceeds the confidence threshold. If the confidence value exceeds the confidence threshold, the DDS 102 may proceed to block 610 and if the confidence value does not exceed the confidence threshold, the DDS 102 may proceed to block 612.

[0083] At block 608, the DDS 102 may perform a set of predefined steps of inspection of the optical film 104 and may further proceed to block 602 to capture further images.

[0084] At block 610, the DDS 102 may detect one or more defects of the optical film 104 by processing the first image.

[0085] At block 612, the DDS 102 may modify one or more motion parameters of the at least one illumination unit 110, the at least one acquisition unit 112 and the optical film 104.

[0086] At block 614, the DDS 102 may receive one or more further images captured by the at least one image acquisition unit 112 based on the modified one or more motion parameters and may redirect to block 604 to recursively iterate blocks 604-614 until the confidence value exceeds the confidence threshold.

[0087] Fig- 7 illustrates an exemplary flowchart of a method for detecting one or more defects of the optical film 104 in accordance with another embodiment of the present disclosure.

[0088] The method 700 may be described in the general context of computer executable instructions. Generally, computer executable instructions can include routines, programs, objects, components, data structures, procedures, modules, and functions, which perform specific functions or implement specific abstract data types. The order in which the method 400 is described is not intended to be construed as a limitation, and any number of the described method blocks can be combined in any order to implement the method. Additionally, individual blocks may be deleted from the methods without departing from the scope of the subject matter described herein. Furthermore, the method can be implemented in any suitable hardware, software, firmware, or combination thereof.

[0089] At block 702, the DDS 102 may receive a first image of the optical film 104 captured by the at least one image acquisition unit 112.

[0090] At block 704, the DDS 102 may classify if the first image is related to a defective film or a defect-free film.

[0091] At block 706, the DDS 102 may determine if the confidence value of the classification exceeds the confidence threshold upon classifying the first image as related to the defective film.

[0092] At block 708, the DDS 102 may modify one or more motion parameters of the at least one illumination unit 110, the at least one acquisition unit 112 and the optical film 104, upon determining that the confidence value does not exceed the confidence threshold.

[0093] At block 710, the DDS 102 may receive one or more further images captured by the at least one acquisition unit based on the modified one or more motion parameters.

[0094] At block 712, the DDS 102 may verify if the confidence value exceeds the confidence threshold. If the confidence value exceeds the confidence threshold, the DDS 102 may proceed to block 714 and if the confidence value does not exceed the confidence threshold, the DDS 102 may proceed to block 706 and may recursively iterate the blocks 706-710 until the confidence value exceeds the confidence threshold. At block 714, the DDS 102 may detect one or more defects of the optical film 104 by processing the first image.

[0095] The illustrated steps are set out to explain the exemplary embodiments shown, and it should be anticipated that ongoing technological development will change the manner in which particular functions are performed. These examples are presented herein for purposes of illustration, and not limitation. Further, the boundaries of the functional building blocks have been arbitrarily defined herein for the convenience of the description. Alternative boundaries can be defined so long as the specified functions and relationships thereof are appropriately performed. Alternatives (including equivalents, extensions, variations, deviations, etc., of those described herein) will be apparent to persons skilled in the relevant art(s) based on the teachings contained herein. Such alternatives fall within the scope of the disclosed embodiments. Also, the words "comprising," "having," "containing," and "including," and other similar forms are intended to be equivalent in meaning and be open ended in that an item or items following any one of these words is not meant to be an exhaustive listing of such item or items or meant to be limited to only the listed item or items. It must also be noted that as used herein and in the appended claims, the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise.

[0096] Finally, the language used in the specification has been principally selected for readability and instructional purposes, and it may not have been selected to delineate or circumscribe the inventive subject matter. Accordingly, the disclosure of the embodiments of the disclosure is intended to be illustrative, but not limiting, of the scope of the disclosure.

[0097] With respect to the use of substantially any plural and / or singular terms herein, those having skill in the art can translate from the plural to the singular and / or from the singular to the plural as is appropriate to the context and / or application. The various singular / plural permutations may be expressly set forth herein for sake of clarity.

Claims

CLAIMS1. A system to detect one or more defects in an optical film, the system comprising: at least one illumination unit movable at one or more angles; at least one image acquisition unit movable at one or more angles; and an optical film mounted at an angle relative to the at least one illumination unit and the at least one image acquisition unit; a processor communicatively coupled with the optical film, the at least one illumination unit and the at least one image acquisition unit; characterized in that, wherein the processor is configured to: receive a first image of the optical film captured by the at least one image acquisition unit; classify the first image as related to one of a defective film or a defect-free film; upon classifying the first image as related to the defective film, determine a confidence value of the classification exceeds a confidence threshold to detect one or more defects related to the defective film; upon determining that the confidence value does not exceed the confidence threshold value, modify one or more motion parameters of one or more of the optical film, the at least one illumination unit and the at least one image acquisition unit to capture one or more further images of the optical film; receive the one or more further images captured by the at least one image acquisition unit based on the modified one or more motion parameters; recursively determine the confidence value of classification of the one or more further images, modify the one or more motion parameters upon determining the confidence value does not exceed the confidence threshold and receive the one or more further images until the confidence value exceeds the confidence threshold; and detect one or more defects in the optical film by processing the one or more further images.

2. The system as claimed in claim 1, wherein the optical film comprises at least a transparent layer, a reflective layer, a metallic coating layer, and an adhesive layer, andwherein the reflective layer comprise a plurality of nano-imprinted structures, like, prism shaped reflective optical structures.

3. The system as claimed in claim 1, wherein the at least one illumination unit is configured to illuminate the optical film using light of one or more wavelengths, and wherein the at least one image acquisition unit is configured to capture one or more images of the illuminated optical film for detecting the one or more defects in the optical film.

4. The system as claimed in claim 1, wherein the at least one image acquisition unit comprises: a first image acquisition unit mounted on a first side of the optical film and configured to capture one or more images of the optical film from the first side of the optical film; and a second image acquisition unit mounted on a second side of the optical film and configured to capture one or more images of the optical film from the second side of the optical film, wherein the second side is an opposite side of the first side.

5. The system as claimed in claim 1, wherein the at least one illumination unit comprising a first illumination unit mounted at an acute angle from an axial plane of the optical film and a second illumination unit mounted perpendicular to an axis of the axial plane of the optical film.

6. The system as claimed in claim 1, wherein the at least one image acquisition unit is configured to capture the first image at an initial position of the at least one image acquisition unit, wherein the initial position may comprise one of an acute angle from an axial plane of the optical film or at an angle perpendicular to the axial plane of the optical film..

7. The system as claimed in claim 1, wherein to classify the first image as related to one of the defective film or the defect-free film, the processor is configured to: compare the first image with a plurality of defective film images and a plurality of defect-free film images; and classify the first image as related to one of:the defect-free film upon determining that a similarity between the first image and at least one of the plurality of defect-free film images exceeds a first threshold value; and the defective film upon determining that a similarity between the first image and at least one of the plurality of defective film images exceeds a second threshold value.

8. The system as claimed in claim 1, wherein the processor is configured to determine the confidence value based on a similarity between the first image and at least one of a defect-free film image or a defective film image.

9. The system as claimed in claim 1, wherein upon classifying the first image as the defect- free film, the processor is configured to: modify one or more motion parameters of one or more of the optical film, the at least one illumination unit and the at least one image acquisition unit to capture one or more further images of the optical film based on a plurality of pre-defined inspection patterns; classify the one or more further images as related to one of a defective film or a defect-free film; and upon classifying the one or more further images as related to the defect-free film, determine that the optical film is successfully inspected; or upon classifying at least one of the one or more further images as related to the defective film, detect one or more defects of the optical film.

10. The system as claimed in claim 1, wherein the one or more motion parameters comprise one or more of: a position and an angle of the at least one acquisition unit and a type of image sensor to capture images; an intensity, a wavelength, a position and an angle of the at least one illumination unit; and a position and an angle of the optical film.

11. The system as claimed in claim 1, wherein to modify the one or more motion parameters, the processor is configured to:identify a class of defect in the first image based on the class of a defective film image, wherein the similarity between the first image and the defective film image exceeds a second threshold value; and modify the one or more motion parameters based on the class, wherein the class is related to one of a transparent layer, a reflective layer, a metallic coating layer, an adhesive layer, the optical film, a function, an environment of the optical film.

12. The system as claimed in claim 1, wherein to detect one or more defects in the optical film by processing the one or more further images, the processor is configured to: pre-process a further image of the one or more further images to remove noise from the further image; process the pre-processed image to identify one or more Regions of Interest (ROIs); compare, using a machine learning (ML) model, each of the one or more ROIs with each of a set of defective film images associated with a class of defect of the first image, wherein the set of defective film images correspond to a plurality of sub-classes of defects; evaluate, using the ML model, a similarity between each of one or more ROIs of the gray-scale image with each of the set of defective film images based on the comparison; compare, using the ML model, the similarity with a similarity threshold value; and upon determining that the similarity of an ROI of the one or more ROIs exceeds the similarity threshold value, classify, using the ML model, the ROI is a defect and the sub-class of the defect corresponding to the defective film image; or upon determining that the similarity of the ROI does not exceed the similarity threshold value, classify, using the ML model, the ROI is new.

13. The system as claimed in claim 12, wherein to train the ML model, the processor is configured to: retrieve a plurality of defective film images and a plurality of defect-free images from a defect database, wherein the defect database is curated by capturing a plurality of images of a plurality of optical films and annotating each of the plurality of images as related to one of a defective film and a defect-free film, wherein annotating theoptical film as the defective film comprises annotating the optical film as comprising one or more classes of defects; and train the ML model to learn one or more image features of the plurality of defective images and the plurality of defect-free images.

14. The system as claimed in claim 1, wherein the processor is further configured to: generate one or more indications of the one or more defects on the corresponding one or more further images; and transmit one or more notifications indicating the one or more indications of the one or more defects on the corresponding one or more further images, one or more classes and sub-classes of defects.

15. The system as claimed in claim 12, wherein upon classifying the ROI is new, the processor is further configured to: notify a user indicating the new ROI related to a defect; and receive one or more inputs from the user related to a class and a sub-class of the defect.

16. The system as claimed in claim 1, further comprising: a rolling unit configured to roll the optical film between a winding roller and an unwinding roller.

17. The system as claimed in claim 16, wherein the processor is configured to modify one or more motion parameters of one or more of the winding roller and the unwinding roller upon determining that the confidence value of classification of the first image does not exceed the confidence threshold.

18. A system to detect one or more defects in an optical film, the system comprising: two image acquisition units configured to capture one or more images of the optical film, wherein a first image acquisition unit mounted on a first side of the optical film and configured to capture one or more images of the optical film from the first side of the optical film and a second image acquisition unit mounted at a second side of the optical film and configured to capture one or more images of the optical film from thesecond side of the optical film, wherein the second side is an opposite side of the optical film; two illumination units configured to illuminate the optical film, wherein a first illumination unit mounted at an acute angle from an axial plane of the optical film and a second illumination unit mounted perpendicular to an axis of the axial plane of the optical film; an image processing unit coupled with the two image acquisition units and configured to process the one or more images to detect one or more defects in the optical film; one or more positioning units coupled with one or more of the two image acquisition units, the two illumination units and the optical film and configured to control a movement of one or more of the optical film, the two image acquisition units, the two illumination units; and a processor coupled with the two image acquisition units, the two illumination units, the one or more positioning units and the image processing unit and configured to: control at least a movement of one or more of the two image acquisition units, the two illumination units, and the optical film to capture one or more images of the optical film; detect one or more defects using a machine learning model by processing the one or more images; mapping the one or more defects on the one or more images; and transmitting one or more notifications related to the one or more defects to a user device.

19. A method for detecting one or more defects in an optical film, the method comprising: receiving a first image of the optical film captured by the at least one image acquisition unit; classifying the first image as related to one of a defective film or a defect-free film; upon classifying the first image as related to the defective film, determining a confidence value of the classification exceeds a confidence threshold value to detect one or more defects related to the defective film;upon determining that the confidence value does not exceed the confidence threshold value, modifying one or more motion parameters of one or more of the optical film, at least one illumination unit and the at least one image acquisition unit to capture one or more further images of the optical film; receiving the one or more further images captured by the at least one image acquisition unit based on the modified one or more motion parameters; recursively determining the confidence value of classification of the one or more further images, modifying the one or more motion parameters upon determining the confidence value does not exceed the confidence threshold and receiving the one or more further images until the confidence value exceeds the confidence threshold; and detecting one or more defects in the optical film by processing the one or more further images.

20. The method as claimed in claim 19, wherein the optical film comprises at least a transparent layer, a reflective layer, a metallic coating layer, and an adhesive layer, and wherein the reflective layer comprise a plurality of prism shaped reflective optical structures.

21. The method as claimed in claim 19, wherein classifying the first image as related to one of the defective film or the defect-free film comprises: comparing the first image with a plurality of defective film images and a plurality of defect-free film images; and classifying the first image as related to one of the defect-free film upon determining that a similarity between the first image and at least one of the plurality of defect-free film images exceeds a first threshold value; and the defective film upon determining that a similarity between the first image and at least one of the plurality of defective film images exceeds a second threshold value.

22. The method as claimed in claim 19, comprises determining the confidence value based on a similarity between the first image and at least one of a defect-free film image or a defective film image.

23. The method as claimed in claim 19, wherein upon classifying the first image as the defect-free film, the method comprises: modifying one or more motion parameters of one or more of the optical film, the at least one illumination unit and the at least one image acquisition unit to capture one or more further images of the optical film based on a plurality of pre-defined inspection patterns; classifying the one or more further images as related to one of a defective film or a defect-free film; and upon classifying the one or more further images as related to the defect-free film, determining that the optical film is successfully inspected; or upon classifying at least one of the one or more further images as related to the defective film, detecting one or more defects of the optical film.

24. The method as claimed in claim 19, wherein the one or more motion parameters comprise one or more of: a position and an angle of the at least one acquisition unit and a type of image sensor to capture images; an intensity, a wavelength, a position and an angle of the at least one illumination unit; and a position and an angle of the optical film.

25. The method as claimed in claim 19, wherein modifying the one or more motion parameters comprises: identifying a class of defect in the first image based on the class of a defective film image, wherein the similarity between the first image and the defective film image exceeds a second threshold value; and modifying the one or more motion parameters based on the class, wherein the class is related to one of a transparent layer, a reflective layer, a metallic coating layer, an adhesive layer, the optical film, a function, an environment of the optical film.

26. The method as claimed in claim 19, wherein detecting one or more defects in the optical film by processing the one or more further images comprises: pre-processing a further image of the one or more further images to remove noise from the further image;processing the pre-processed image to identify one or more Regions of Interest(ROIs); comparing, using a machine learning (ML) model, each of the one or more ROIs with each of a set of defective film images associated with a class of defect of the first image, wherein the set of defective film images correspond to a plurality of sub-classes of defects; evaluating, using the ML model, a similarity between each of one or more ROIs of the gray-scale image with each of the set of defective film images based on the comparison; comparing, using the ML model, the similarity with a similarity threshold value; and upon determining that the similarity of an ROI of the one or more ROIs exceeds the similarity threshold value, classifying, using the ML model, the ROI is a defect and the sub-class of the defect corresponding to the defective film image; or upon determining that the similarity of the ROI does not exceed the similarity threshold value, classifying, using the ML model, the ROI is new.

27. The method as claimed in claim 26, the method comprises training the ML model by: retrieving a plurality of defective film images and a plurality of defect-free images from a defect database, wherein the defect database is curated by capturing a plurality of images of a plurality of optical films and annotating each of the plurality of images as related to one of a defective film and a defect-free film, wherein annotating the optical film as the defective film comprises annotating the optical film as comprising one or more classes of defects; and training the ML model to learn one or more image features of the plurality of defective images and the plurality of defect-free images.

28. The method as claimed in claim 19, further comprises: generating one or more indications of the one or more defects on the corresponding one or more further images; and transmitting one or more notifications indicating the one or more indications of the one or more defects on the corresponding one or more further images, one or more classes and sub-classes of defects.

29. The method as claimed in claim 27, wherein upon classifying the ROI is new, the method comprises: notifying a user indicating the new ROI related to a defect; and receiving one or more inputs from the user related to a class and a sub-class of the defect.

Images