TWI463418B - Monolithic image perception device and method - Google Patents

Description

Single stone image sensing device and method

The present invention generally relates to imaging devices. In particular, the present invention relates to microdevices for image recognition disposed on a transparent substrate such as glass or embedded in a transparent substrate such as glass.

Transparent surfaces such as glass have been around for hundreds of years. The transparent surface was originally intended to protect the living space while giving the occupants a sense of the outside world (landscape, weather and possible threats). More recently, a large number of transparent surfaces have been required for the display industry, starting with cathode ray tubes (CRTs) and more recently for liquid crystal displays (LCDs) and many other types of flat panel displays. In use, in most cases, humans or living organisms (animals, plants) are located near such transparent surfaces.

Image sensors have been in use for decades (for example, CCD or CMOS sensors). For example, see U.S. Patent No. 6,617,565 for a single-chip CMOS image sensor, the contents of which are incorporated herein by reference. A typical image sensor is based on a camera design and generally includes an integrated circuit located behind the lens, which may be micro or movable (eg, a spiral mounted lens). The sensor is used to convert light energy (photons) into electrical signals that are proportional to the amount of light received by the photosensitive elements organized in an array on the inductor. The image is synthesized based on the output of the photosensitive element.

Image recognition technology has become increasingly needed. Cameras of all sizes and manufacturing are needed for applications such as security, identification, intelligence, quality inspection, traffic surveillance and more. The camera is typically linked to the display device by a wired or wireless connection. Today, mobile phones are conventionally equipped with a miniature camera that is connected to an LCD display device disposed therein.

Advanced image recognition requires high-resolution imaging synthesis. Today's image recognition systems operate at relatively low speeds due to insufficient processing power and/or because the processor can only process one pixel of the image at a time.

Therefore, there is a need for a new imaging recognition device that has been improved in the prior art.

It is an object of the present invention to provide an image recognition device having a sensing region (e.g., a light sensitive element) that is directly included in a transparent or translucent material that forms an optical interface between an incident image and a sensing region. The image recognition device itself is preferably transparent or translucent.

Yet another object of the present invention is to provide a sensing zone with "regional" decision making capabilities by means of an array of trainable processing elements. In one embodiment of the invention, a trainable cognitive memory element or unit is associated with one or more photosensitive elements. The regional decision making capability provides the advantage of reducing the transmission requirements (i.e., bandwidth) of the device, especially when the number of photosensitive elements is large and when the transmission frequency of the photosensitive elements must be high. High resolution, high speed imaging devices can be obtained by providing a larger array of sensing regions each having regional decision making capabilities.

According to embodiments of the present invention, the trainable cognitive memory elements can operate in parallel at low frequencies and cause very low currents. Therefore, the self-discipline operation of each component is ensured, and a very economical energy source such as a solar cell or an equalizer can be used.

In accordance with an embodiment of the present invention, a novel solitolite image recognition device is formed from all of one or more light sensitive elements embedded in a substrate associated with one or more trainable cognitive memory elements.

In accordance with embodiments of the present invention, a plurality of photosensitive elements associated with a plurality of trainable cognitive elements can be configured in one or more arrays and spread over a flat transparent or translucent substrate. The array can have variable geometry and connectivity. A typical geometry may be, but is not limited to, a linear array of parallel neurons, or a two-dimensional array of neurons connected in a rectangular matrix or honeycomb geometry.

Further applications and advantages of various embodiments of the present invention are discussed below with reference to the drawings.

While the invention may be embodied in a variety of different forms, the present invention is described herein, and it is understood that / or any particular preferred embodiment illustrated.

The present invention is an imaging device that can include an inductor sensing device, such as a light sensitive element, coupled to a trainable cognitive element, in conjunction with or otherwise associated with the trainable cognitive element Correspondingly, two of the elements are chemically deposited or otherwise deposited on the surface of the transparent substrate or embedded in the surface of the transparent substrate. Throughout this document, the combination of a sensing zone with a trainable cognitive component with "regional" decision making capabilities is referred to as a "cognitive sensor." Throughout this document, the trainable cognitive component is called "CogniMem". The sensing area typically consists of one or more photosensitive elements, but may encompass other sensing configurations.

In accordance with embodiments of the present invention, a cognitive sensor can be configured to recognize an incident light pattern (eg, an image or image portion), process an incident light pattern to produce an indication of regional decisions and transmission region decisions. Cognitive sensors can include many components such as, but not limited to, regional decision making capabilities - data entry logic, "neurons" and decision output logic, memory buffers, solar cells for energy autonomy, and more. many. Each cognitive sensor preferably characterizes a reactive associative learning memory (REALM) in a parallel configuration. In accordance with an embodiment of the present invention, CogniMem is capable of digital or analog pattern recognition without any computer instructions.

A CogniMem can contain one or more neurons that are parallel-accessible related memories that can react to input patterns that are similar in their own content. Neurons can react individually or collectively by enhancing their response based on responses from other neighboring neurons. This selection can be done via a suppression/excitation input line connected to the neuron.

The neuronal content of CogniMem constitutes "knowledge." Knowledge is a set of static distinct digital signatures. Knowledge can be static knowledge (one-time load) or dynamic knowledge (updated by other neurons' responses or properly loaded from an external knowledge base), but is preferably automatically generated by the learning process without the need for a computer. do. The same or different knowledge can be used for CogniMem deposited on the same substrate.

The CogniMem can be deposited on the substrate or embedded in the substrate (or otherwise coupled to the substrate) as part of the cognitive sensor, or as a separate part. In the former case, CogniMem is usually dedicated to identifying the pixel data transmitted by the photosensitive element. In the latter case, CogniMem can be used to support other CogniMems and can be used, for example, to identify different data types transmitted by other CogniMem units (eg, to incorporate response patterns from multiple cognitive sensors).

The following is a list of patents and published applications, the entire contents of each of which are hereby incorporated by reference, for each of the entire disclosures of the disclosure of the disclosure of the disclosure of the disclosure of the disclosure of the disclosure of the disclosures of Neuron circuit; modified neuron circuit architecture No. 5,717,832; circuit 5,701,397 for precharging idle neuron circuits; karaoke circuit for series connection of neuron circuits; 5,740,326 for search/ Circuits for classifying data in neural networks; No. 6,332,137 for parallel-related memory for separate hardware identification; single-line search and classification for No. 6,606,614; Japanese application JP8-171543 for daisy-chaining of series connection of neuron circuits Circuit; JP8-171542 advanced load circuit; JP8-171541 aggregation circuit (search/classification); JP8-171540 neural network and neural chip; JP8-069445 neuron circuit architecture; Korean patent application KR164943 innovative neuron circuit architecture; Patent EP0694852 innovative neuron circuit architecture; EP0694854 improved neural semiconductor chip architecture; EP0694855 neural network search / classification; EP0694853 for pre-charging the input vector component of the idle neuron circuit in the identification phase; EP0694856 for the serial connection of the neuron circuit daisy chain circuit; Canadian application CA2149478 modified neuron circuit Architecture; Canadian patent CA2149479 improves the neural semiconductor wafer architecture.

The number of neurons constructed on CogniMem can be changed from 1 to N, which is theoretically infinite due to the architecture of the neuronal unit. Currently, N can be as high as about 1000. In general, N is determined by the application, and in particular, N is determined by the diversity of patterns to be recognized and the type of decision being transmitted. Those skilled in the art will recognize that sputum technology can be an important factor in determining the number of neurons that can be provided per unit area.

1A and 1B illustrate an exemplary configuration of an image recognition device in accordance with an embodiment of the present invention. 1A is a top plan view of a device 100 that includes a substrate 102 that can be made from a variety of transparent or translucent materials such as glass, plastic glass, transparent plastic, and the like. One or more cognitive sensors 104 (in this case, in an array) can be embedded in the substrate 102 or attached or glued to the surface of the substrate 102, or otherwise The surface of 102 is coupled (see Figure 1B). An optical path can be etched or deposited on the substrate before each photosensitive element. For example, etched at the location of the cognitive sensor 104 of the substrate 102 to create a lens 102a for each cognitive sensor 104. Alternatively, the microlens 102a can be inserted into the substrate 102 (Fig. 2) or glued to (Fig. 3A-3B) the substrate 102 in front of the photosensitive element. Another option may be to change the substrate to change the refractive index of the portion of the substrate adjacent each inductor to focus the incident light. As shown in FIG. 1B, incident light is collected by the substrate lens 102a on each of the cognitive sensors 104.

The plurality of lenses 102a allow the cognitive sensor 104 to encompass a variety of fields of view, preferably equal to the surface of the substrate, but may also cover a field of view that is narrower or greater than the field of view of the surface of the substrate. The microlens 102a changes the array of cognitive sensors 104 into a telecentric image sensing device with an unrestricted surface and field of view.

2 is a top plan view of a single stone imaging apparatus in accordance with another embodiment of the present invention. As shown, the lens 102a is embedded in the substrate 102 and disposed over each of the cognitive sensors 104. As an example of use of the imaging device, the display DNA fragment 202 is disposed on the surface of the substrate 102. Each cognitive sensor 104 can be configured to recognize a particular DNA segment individually or in cooperation with a neighboring cognitive sensor 104 and output a signal when the segment has been identified.

3A-3B illustrate an exemplary embodiment of an individual cognitive sensor 104. As shown in FIG. 3A, the concentrated neuron region 104a surrounds the pixel sensing region 104b. Neurons in neuron region 104a can be coupled to sensing elements in pixel region 104b and can be configured to recognize patterns induced by pixel region 104b. As shown in FIG. 3B, a convex lens or microlens 102a is disposed over the pixel region 104b on the surface of the substrate 102 for focusing incident light on the pixel region 104b or directly connected to the inductor without an intermediate substrate. Lens 102a can be chemically deposited on a substrate, for example, by conventional methods.

4 is a functional block diagram of an exemplary cognitive sensor 104 in accordance with an embodiment of the present invention. The cognitive sensor 104 includes a sensor or sensing region 402, data presentation logic 404, neural network 406, and region decision logic 408. The sensor 402 can include one or more inductive elements (such as light sensitive elements). Data presentation logic 404 is coupled to sensing area 402 and neural network 406 and is configured to present data output from the sensor to the neurons in a manner suitable for processing. Neuron 406 is or becomes subject to knowledge "teaching" and can process the data input from presentation logic 404 to neuron 406 and output the processed data to regional decision logic 408 based on the processing data generation decision. The region decision logic 408 can be coupled to other cognitive sensors or CogniMems by a variety of known methods. Thus, the cognitive sensor 104 can be configured in an array of arrays and arrays.

5A and 5B show the configuration of an array of cognitive sensors. As shown in FIG. 5A, each cognitive sensor 104 can be coupled to a plurality of cognitive sensors 104 for use in array 502. As described below, the input and output busses can be used for series or parallel coupling of inductors.

As shown in FIG. 5B, each array 502 can be coupled to a plurality of arrays 502 to form an array of groups 504. By configuring an array of arrays of cognitive sensors 104, a highly efficient identification device with high resolution and high speed is produced. That is, the resolution of the imaging device can be increased by increasing the number of sensors. However, by providing a robust regional decision making capability in the form of CogniMem, the increase in the number of cognitive sensors does not reduce the processing speed of the device. Moreover, it should be understood that the array can be organized in a number of different geometries, and the invention is not limited to square arrays.

As mentioned above, each neuron can be coupled to a plurality of inputs 1-n, which can be, for example, a multiplex input, but are not limited thereto. Figure 6A is a representation of a neuron having multiple inputs, which is simplified in Figure 6B. Thus, the array of neurons can be combined using input bus 602 (the bus bar 602 is absent on Figure 6C), as shown in the simple parallel architecture of Figure 6C. Each output of neuron 406 can be coupled to global decision bus 406.

7 is a functional block diagram of an exemplary neuron in accordance with an embodiment of the present invention. The purpose of organizing neurons that are infinitely expanding networks is to learn and recall digital vectors or signatures (patterns). The digital signature is primarily a spatial distribution of the intensity of light encoded by the data reduction process. The neurons can be connected in parallel as represented in Figure 6C, which means that all neuron inputs and all their outputs are connected in parallel.

The data signal can be input to the neuron 700 from the multiplex input bus (not shown). The learning cognitive multiplexer 702 can divide the multiplexed input signals and transmit the input data signals to the neuron recall memory 704 and associated logic elements 706. The neuron recall memory 704 processes the input signal and outputs the processed signal to the associated logic element 706. Correlation logic element 706 includes a similarity coefficient decision element 706a.

Each neuron can receive a broadcast pattern (i.e., a vector representing the digital signature of the sensor data) generated by the data presentation logic 404. This broadcast pattern can be used to convert data (simplified data) for sensors generated instantaneously or in the time domain.

Neurons have three possible subsequent timing states: sleep, forthcoming learning (RTL), and subsequent determination. At least one neuron is in the RTL state at all times unless the network is full (ie, all neurons will be determined). If all of the parallel connected neurons are considered to be a chain, the RTL neurons can move from the first position of the chain to the last position. In the case indicated here, the RTL neurons will typically be on the right side of the determined neurons and the dormant neurons will be on the right side of the RTL neurons.

When a neuron is dormant, it will not react to any incident pattern. The RTL neuron will load the incident pattern into its recall memory to learn it during the user process decision learning. This RTL neuron will not participate in the identification process, but will be dedicated to learning new knowledge.

The learning process involves generating new knowledge when an unknown pattern appears and the user decides to learn it. This knowledge addition will occur in RTL neurons. In addition to generating new knowledge, certain neurons that may incorrectly identify the incident pattern (ie, fail to correlate with the appropriate category) will reduce their similar domains to avoid further misclassification. This situation causes knowledge modification or "adaptive learning."

The optical component can output a digitized radiation measurement. All values above the spatial distribution combine to form a pattern; this pattern can also develop in the time domain and produce a pattern stream. This pattern undergoes a data simplification process that results in a digital signature (vector) of the radiation measurement pattern. The simplification process cannot exceed the so-called "minimum discrimination matrix" described below. To enumerate examples of 5x7 matrices, it is feasible to distinguish all European capital letters instead of Chinese characters, which requires a 16x16 matrix.

When it is determined that the neuron is in the RTL state, it learns the pattern by associating the vector loaded into the recall memory 704 with the category stored in the class register 709. When the incident pattern enters the determining neuron, the learning/recognition multiplexer 702 will transmit it to the correlation logic 706 to estimate the similarity of this pattern to the vector stored in the recall memory 704. If the calculated similarity is found to be less than or equal to the similarity coefficient 706a, then the neuron will be energized, and thus the signal passes through logic 712. The function of the stimulus/suppression logic is to perform global arbitration among all defined "aggressive" (ie, excited) neurons when many neurons become excited, and "suppress" those that do not have the best similarity Neurons.

Area of concern

Each cognitive sensor can be associated with a region of interest (ROI) for a video frame. Each cognitive sensor can extract the signature of the ROI for broadcast to its neurons (for learning or identification purposes). The signature of the ROI is a compressed format of its pixel values that is reduced to fit a sequence of N values, where N is the size of the neuron memory unit.

An example of a memory capacity with 256-bit tuples is listed. The cognitive sensor classifies the rectangular ROI of NxM pixels. The ROI signature will be reduced to 256 values from NxM values by, for example, simple block compression.

The cognitive sensor can be configured to handle ROI of any shape, and the signature extraction can be selected for special applications (eg, partial inspection, surface inspection, facial recognition, target tracking, etc.). Some signature extractions can integrate time, repeatability, and more. Also, the neuron can be equipped with a memory unit greater than 8 bits to supply a pixel resolution of 12 bits or greater to the input of the sensor.

The combination of neurons, together with the sensor and data presentation logic, constitutes a completely novel method for embedding image recognition without any software required for the learning or identification process.

The location of the CogniMem can be either transitive or selective (such as driven by responses from other CogniMem units).

It should be understood that the substrate carrying the cognitive sensor acts as a mechanical support and acts as a lens (see, for example, Figures 1-2). The substrate can be, but is not limited to, a rigid or flexible, flat or curved surface made of glass, plastic glass, plastic, mylar or other materials.

The connection between the cognitive sensor and the CogniMem unit on the same substrate should preferably use a minimum number of wires.

The knowledge loaded into the cognitive sensor can better resolve the identification of different series of patterns that are related or unrelated.

Instance

According to an embodiment of the invention, the cognitive sensor is ideal for performing an inspection during the automated manufacturing process. As shown in Figure 8, one or more cognitive sensors can be used to inspect the water bottle. In this example, three different cognitive sensors are used to examine three different regions, referred to as Experts 1-3. The global response can be determined by a combination of three "expert" cognitive sensors.

In this example, cognitive sensor 1 (Expert 1) can be trained to classify the signature of the ROI containing cap 802. Cognitive Sensor 1 can classify its ROI into two categories: good and bad. A bad category can combine several conditions: the cover is missing or the cover is not properly tightened.

Likewise, cognitive sensor 2 (Expert 2) can learn the signature of the ROI that intersects the fluid line 804 in the bottle. The ROI can be a narrower vertical rectangle and will ideally cover the minimum and maximum possible fill height in the bottle. Depending on the quality control standards of the manufacturer, the cognitive sensor 2 can classify its ROI into any number of categories (for example): acceptable and unacceptable; too high, acceptable and too low; or too high, high but Acceptable; range, low but acceptable, too low.

The cognitive sensor 3 (Expert 3) can learn to cover the signature of the area of interest of the tag area 806. The cognitive sensor 3 can be trained to recognize a variety of conditions or a combination of conditions, such as, for example, missing tags, defective tags (tear, scratches or folds), misplaced tags (reversed, tilted), and Ok.

The output from the cognitive sensors 1-3 can be provided to a controller associated with the automated manufacturing process to take appropriate action based on the decisions generated thereby.

In accordance with embodiments of the present invention, the cognitive sensor can be individually packaged to form a smart photocell or smart microlens. This device is used in many techniques and can be used, for example, to detect moving parts, identify routes during mechanical assembly or route moving parts (Fig. 9A); for biometric identification, such as in camera phones (Fig. 9B) Or for guest detection and identification in door peepholes and the like (Fig. 9C).

In accordance with another embodiment of the present invention, a driver awareness detection system is provided. Referring to Figure 10, one or more of the cognitive sensors 104 can be embedded in a windshield, dashboard flat panel display or headlight of a motor vehicle. The cognitive sensor 104 can be taught to recognize a pattern indicating when the driver is no longer attentive (eg, the driver falls asleep) and output a signal to trigger an alarm. Such patterns may include gaze tracking, facial recognition, facial expression recognition and more. In addition, the cognitive sensor 104 in the windshield or headlight can be taught to identify objects or events external to the vehicle, such as identifying a raindrop with a windshield wiper system, or identifying a driving accident for a traffic accident warning system.

There are many ways to detect objects that appear randomly in the far field or near field of view. For example, two or three sensors can be equipped with lenses that are focused at different distances. Sensors can load the same knowledge, but work with areas of different sizes. If at least one sensor recognizes the object, then the global response of the identification system is considered correct.

Also, cognitive sensors can be designed with input sensors that are sensitive to different wavelengths, such as near IR, IR, color filtered, and the like. For a particular object or scene, these cognitive sensors will produce different pixel values, but can be trained on their individual video images to identify the category of the object. In target tracking, the combination of near-IR and IR cognitive sensors will provide the ability to identify targets at any time of the day.

In accordance with another embodiment of the present invention, an array of cognitive sensors can be used in many other manufacturing applications. For example, as shown in FIG. 11A, a one-dimensional array of cognitive sensors 1102 can be used to inspect a glass float 1103 during manufacturing. As shown in FIG. 11B, a two-dimensional array of cognitive sensors 1104 can be used to detect contaminants at the bottom of the container 1105, such as a beverage bottle. In such applications, each cognitive sensor can be taught to identify a pattern indicative of a glass crack or fluid contaminant.

According to another embodiment of the present invention, a cognitive sensor can be disposed on a glass plane or the like to perform a plurality of independent functions. Cognitive sensors can be grouped and each group taught with different knowledge. 12 shows an example of a sliding glass door 1202 that includes a plurality of cognitive sensor groups 1204 for detecting objects of different sizes. The first group can be taught by knowledge of a first size 1208 that identifies a person or animal (eg, a dog) while the second group can be taught by people of different sizes (eg, boys) 1210, and the third group is used for people of another size (for example, adult) 1212, and so on. Each group of 1204 can be coupled to one or more CogniMem 1206 for control of the sliding door.

It will be apparent to those skilled in the art that the imaging device of the present invention may be useful in many other applications not listed herein, as after reading this patent document. Another application, for example, includes detection of permanent damage (texture changes) in dams, bridges, or other man-made structures. The implementation of this application should be apparent from the above description of embodiments of the invention. In addition, power and signal transmission can be wireless (eg, infrared, photocell, induction coil, etc.).

Thus, many preferred embodiments have been described above in detail with reference to the drawings. Although the present invention has been described in terms of the preferred embodiments thereof, it is apparent to those skilled in the art that the invention may be modified, modified, and substituted in the spirit and scope of the invention.

100. . . Device

102. . . Substrate

102a. . . lens

104. . . Cognitive sensor

104a. . . Neuronal area

104b. . . Pixel area

202. . . DNA fragment

402. . . Sensor/sensing area

404. . . Data presentation logic

406. . . Neural network

408. . . Regional decision logic

502. . . Array

504. . . Array group

700. . . Neurons

702. . . Learning cognitive multiplexer

704. . . Neuron recall memory

706. . . Related logic components

706a. . . Similarity coefficient decision component

709. . . Category register

712. . . logic

802. . . Cap

804. . . Fluid line in the bottle

806. . . Label area

1102. . . One-dimensional array of cognitive sensors

1103. . . Glass float

1104. . . Two-dimensional array of cognitive sensors

1105. . . container

1202. . . Sliding glass door

1204. . . Cognitive sensor group

1206. . . CogniMem

1208. . . First size person or animal

1210. . . People of different sizes

1212. . . Another size person

1A-1B respectively include front and top views of an array of inductors disposed on glass or plastic glass or other transparent plastic or transparent substrate having an etched lens in the glass or plastic glass or substrate, in accordance with an embodiment of the present invention; 2 is a top plan view of an array of inductors disposed on a glass or plexus substrate, showing a detected DNA fragment having an etched lens in the glass or substrate, according to an embodiment of the invention; FIGS. 3A-3B illustrate Side view and top view of an inductor mold of one embodiment of the invention; FIG. 4 is a block diagram of an inductor according to an embodiment of the present invention; FIG. 5A is a block diagram of an inductor array according to an embodiment of the present invention; a block diagram of a sensor array set in accordance with an embodiment of the present invention; FIGS. 6A-6C illustrate a neural configuration in accordance with an embodiment of the present invention; FIG. 7 is a block diagram of a neuron according to an embodiment of the present invention; 8- Figure 12 illustrates an illustrative application of an image recognition device in accordance with an embodiment of the present invention.

100. . . Device

102. . . Substrate

102a. . . lens

104. . . Cognitive sensor

Claims (35)

An image recognition device comprising: a sensing region embedded in a transparent or translucent substrate or disposed on a transparent or translucent substrate; and a processing component coupled to the sensing region, the processing component being embedded Or being disposed on the substrate; wherein the transparent or translucent substrate forms an optical interface between the incident image to be sensed and one of the sensing pixels of the sensing region. The image recognition device of claim 1, wherein the sensing area is transparent or translucent. The image recognition device of claim 1, wherein the substrate comprises glass, plastic glass or other transparent material. The image recognition device of claim 1, wherein the sensing region comprises one or more photosensitive elements, and the processing element comprises one or more cognitive memory elements, wherein each of the photosensitive elements is configured to be based on an input The received light outputs a signal, and each of the cognitive memory elements is configured to recognize a pattern based on the signals output by the photosensitive elements. The image recognition device of claim 4, wherein the cognitive memory component is trainable. The image recognition device of claim 4, wherein each of the cognitive memory elements comprises a multiplex input bus coupled to an input side thereof and coupled to an output side of an output bus. a plurality of neurons, each of which is taught by a knowledge that allows the corresponding neuron Identify a signal and perform a decision. The image recognition device of claim 1, wherein the substrate comprises a plurality of lens portions, each lens portion providing an optical interface with an inductive pixel or pixel region of the image recognition device. The image recognition device of claim 7, wherein each lens portion is formed by etching of the substrate. The image recognition device of claim 1, wherein the image recognition device comprises a plurality of the sensing regions organized in an array. The image recognition device of claim 1, wherein the image recognition device is configured to output a wireless output signal. The image recognition device of claim 1, wherein the device is wirelessly powered. An image recognition device comprising: a plurality of cognitive sensors disposed on a transparent or translucent substrate, each sensor comprising: a photosensitive element; and a trainable cognitive memory unit associated with the photosensitive element And a plurality of optical interfaces formed on the substrate, each of which is optically coupled to a respective one of the plurality of cognitive sensors. The image recognition device of claim 12, wherein the optical interfaces are lenses formed by etching the substrate at a position adjacent to each of the cognitive sensors. The image recognition device of claim 12, wherein each cognitive sensor is Trained and configured to recognize various patterns based on incident light. The image recognition device of claim 12, wherein the substrate comprises glass, plastic glass or other transparent material. The image recognition device of claim 12, wherein each of the cognitive memory elements comprises a multiplex input bus coupled to an input side thereof and coupled to an output side of an output bus. A plurality of neurons, each of which is taught by a knowledge that allows the corresponding neuron to recognize a signal and perform a decision. The image recognition device of claim 16, wherein each cognitive memory unit is taught to identify a different portion of an image, and the plurality of cognitive memory units are configured to operate in common to identify the image. The image recognition device of claim 17, wherein the cognitive sensors are configured to output a signal for the identification of the image. The image recognition device of claim 12, wherein the cognitive sensors operate in parallel at a low frequency. The image recognition device of claim 12, wherein the cognitive sensors operate at a very low current. The image recognition device of claim 12, wherein each cognitive sensor is configured to transmit and receive a plurality of wireless signals. The image recognition device of claim 12, wherein the device is wirelessly powered. An image recognition device comprising: a transparent or translucent substrate; a cognitive sensing member for sensing various patterns of incident light and based on The sensing patterns output a signal, the cognitive sensing member is embedded in the substrate, and the optical interface member is configured to provide an optical interface to the cognitive sensing member; wherein the cognitive sensing member comprises a plurality of cognitive sensors, each The method comprises: a photosensitive element; and a trainable cognitive memory unit associated with the photosensitive element. The image recognition device of claim 23, wherein each of the photosensitive elements is a photodetector, wherein each of the photodetectors is configured to output a signal based on light received at an input, and each The cognitive memory unit is configured to recognize a pattern based on the signals output by the photodetectors. The image recognition device of claim 23, wherein the optical interface member comprises a plurality of lenses formed in the substrate. The image recognition device of claim 23, wherein each of the cognitive memory units comprises a multiplex input bus coupled to an input side thereof and coupled to an output side of an output bus. A plurality of neurons, each of which is taught by a knowledge that allows the corresponding neuron to recognize a signal and perform a decision. The image recognition device of claim 23, wherein the cognitive sensing component performs image recognition operations by digitizing a plurality of parallel components without a software program, each parallel component having self-sufficient, self-regulating behavior. The image recognition device of claim 23, wherein the cognitive sensing component is configured to transmit and receive a plurality of wireless signals. The image recognition device of claim 23, wherein the device is wirelessly powered. An image recognition device comprising a combination of a neuron having a sensor and a data presentation logic, wherein the combination is embedded in a transparent or translucent substrate or disposed on a transparent or translucent substrate. An image recognition method comprising the steps of: embedding a cognitive sensing member in a transparent or translucent substrate or depositing on a transparent or translucent substrate, the cognitive sensing member for sensing a pattern of incident light and based on the The sensing pattern outputs a signal; and provides an optical interface member for providing an optical interface to the cognitive sensing member; wherein the cognitive sensing member comprises a plurality of cognitive sensors, each comprising: a photosensitive element; and a A cognitive memory unit can be trained that is associated with the photosensitive element. The image recognition method of claim 31, wherein each of the photosensitive elements is a photodetector, wherein each of the photodetectors is configured to output a signal based on light received at an input, and each of the The cognitive memory unit is configured to recognize a pattern from the signals output by the photodetectors. The image recognition method of claim 31, further comprising providing the image formation A step of the plurality of lenses in the substrate or deposited on the substrate as part of the optical interface member. The image recognition method of claim 31, wherein each of the cognitive memory units comprises a plurality of input buffers coupled to one of the input sides and coupled to an output side of an output bus. Neurons, each of which is taught by a knowledge that allows the corresponding neuron to recognize a signal and perform a decision. An image recognition method comprising: using a plurality of optical interfaces to provide an optical path to one of a plurality of sensing elements, the plurality of optical interfaces being embedded in a transparent or translucent substrate or provided on a transparent or translucent substrate; The plurality of sensing elements are embedded in or provided on the substrate; and a plurality of signals generated from the plurality of sensing elements are processed in parallel in the plurality of processing elements, each of the plurality of processing elements being coupled to the plurality of processing elements One of a plurality of sensing elements and each of the plurality of processing elements is embedded in or provided on the substrate.