CN117635678A

CN117635678A - Light intensity information extraction method of regional spectrum chip

Info

Publication number: CN117635678A
Application number: CN202210984828.XA
Authority: CN
Inventors: 方勇; 张鸿; 黄志雷; 王宇
Original assignee: Beijing Heguang Technology Co ltd
Current assignee: Beijing Heguang Technology Co ltd
Priority date: 2022-08-17
Filing date: 2022-08-17
Publication date: 2024-03-01

Abstract

The invention provides a light intensity information extraction method of a split-region spectrum chip, wherein the light intensity extraction method comprises the following steps of: determining the texture direction of an object to be identified; determining a related common pixel set according to the texture direction; and optimizing the measured value of the corresponding structural pixel based on the measured value of the selected common pixel set to obtain the optimized measured value of the structural pixel.

Description

Light intensity information extraction method of regional spectrum chip

Technical Field

The invention relates to the technical field of spectrum information, in particular to a light intensity information extraction method of a split-region spectrum chip.

Background

Along with the development of science and technology and economy, imaging modules are increasingly used, such as photography, image recognition, machine vision, and the like. The sensor for the imaging module at present mainly comprises a CMOS chip, a CCD chip and the like, and can be used for acquiring RGB images, and the traditional imaging module has inherent defects, so that on one hand, the traditional imaging module covers a visible light wave band and is easily influenced by external visible light, on the other hand, the imaging module is lack of information of spectrum details because of only having R, G, B channels, that is to say, the imaging module can not truly reflect the information of incident light, and deviation easily occurs during imaging.

The spectrum is a fingerprint of a substance, and the spectrum information contains energy information of incident light at each wavelength, so that compared with the existing imaging module, the spectrum information contains more information and is more accurate. Therefore, the spectrum module based on the spectrum chip gradually appears in the field of vision of people, how to effectively and efficiently use the spectrum chip to acquire more accurate information becomes a problem to be solved.

With the development of computer technology, a new spectrometer type has emerged in recent years: a reconstruction spectrometer is calculated which approximates or even reconstructs the spectrum of the incident light by calculation. The computational reconstruction spectrometer can relatively better solve the problem of detection performance degradation caused by miniaturization. The fingerprint is photographed through the spectrum chip, so that a clearer image is required, and spectrum information is required to be obtained. Because the object to be identified is not necessarily regular, a certain texture can exist, at the moment, after the light source irradiates the object to be identified, incident light is generated to enter the spectrum chip, and the result is inaccurate due to the texture direction of the fingerprint.

Disclosure of Invention

The invention provides a light intensity information extraction method of a regional spectrum chip, wherein the light intensity information extraction method is used for identifying textures, selecting corresponding common pixels, averaging measured values based on the common pixels, and optimizing the measured values of the corresponding structural pixels based on the average values, so that the identification precision or the spectrum recovery precision is improved.

The invention further provides a light intensity information extraction method of the split-region spectrum chip, wherein the light intensity information extraction method firstly determines the texture direction, namely when the structural pixels are positioned in the valley region, the peripheral common pixel set positioned in the valley region can be determined through the texture direction, the corresponding structural pixels correspond to the ridge region, and the peripheral common pixel set corresponding to the ridge region can also be taken.

The invention further provides a light intensity information extraction method of the split-region spectrum chip, wherein the method can be based on a gradual change method, can also be used for carrying out preliminary recovery on a fingerprint texture image, and is beneficial to identifying the texture direction of the fingerprint by carrying out texture judgment according to the texture image.

The invention further provides a light intensity information extraction method of the split-region spectrum chip, wherein the light intensity information extraction method optimizes the measured value of the corresponding structural pixel through the measured value of the selected common pixel set, and then identifies whether the object is a living body or not according to the optimized measured value, or recovers the spectrum curve according to the optimized measured value.

The invention further provides a light intensity information extraction method of the split-region spectrum chip, wherein the imaging quality can be improved based on auxiliary imaging of optimized measured values; because the spectrum chip partial area has the structural pixels, the measured values of the structural pixels and the common pixels can be greatly different, and the image can be restored based on the optimized measured values, so that the influence of the structural pixels on imaging is compensated.

According to one aspect of the present invention, the foregoing and other objects and advantages are achieved by a light intensity information extraction method of a split-region spectrum chip of the present invention, wherein the light intensity extraction method comprises:

step S1, determining the texture direction of an object to be identified;

step S2, determining a related common pixel set according to the texture direction; and

and step S3, optimizing the measured value of the corresponding structural pixel based on the determined measured value of the common pixel set to obtain the optimized measured value of the structural pixel.

According to one embodiment of the present invention, in the step S1, a gradient algorithm is adopted to calculate a local brightness gradient of the spectrum chip, and a direction perpendicular to the gradient direction is a texture direction.

According to one embodiment of the present invention, in the step S1, the object to be identified is imaged by the spectrum chip, a texture image is obtained, and the texture direction is identified based on the texture image.

According to an embodiment of the present invention, step S2 in the light intensity information extraction method further includes:

step S2.1, determining two sampling points according to the set sampling distance and the texture direction; and

and step S2.2, determining common pixels in the field according to the sampling points and forming the common pixel set by the sampling points.

and S2.3, removing common pixels with excessive difference of measured values in the common pixel set, and removing structural pixels adjacent to the sampling points.

According to an embodiment of the present invention, the step S3 in the light intensity information extraction method further includes:

step S3.1, processing a measured value of a common pixel set by a averaging method and obtaining a processing value corresponding to the structural pixel; and

and step S3.2, optimizing the measured value of the structural pixel according to the processing value, and obtaining an optimized measured value.

According to one embodiment of the invention, the optimized measurement of the structure pixel is the measurement of the structure pixel minus the averaged treatment value or is the ratio of the measurement of the structure pixel to the averaged treatment value.

s2.4, selecting a plurality of common pixels around the structural pixel to form a common pixel set; and

s2.5, determining the weight of each common pixel point according to the distance between the center point of the structural pixel and each common pixel center point and the included angle theta between the line segment formed by the two points and the gradient direction.

s2.6, setting a threshold value x, and selecting common pixels with weight equal to or greater than x as a common pixel set.

According to an embodiment of the present invention, the step S3 of the light intensity information extraction method further includes:

step S3.3, obtaining a processing value according to the measured value of each common pixel and the weight thereof; and

and step S3.4, optimizing the measured value of the structural pixel according to the processing value, and obtaining an optimized measured value.

According to an embodiment of the present invention, the light intensity information extraction method further includes:

step S4 is based on the obtained optimized pixel values of the structural pixels, corresponding to the incident spectrum f (λ).

According to one embodiment of the invention, the corresponding incident spectrum f (λ) is calculated using the optimized measurement values and the corresponding transmission spectrum matrix; or recovering the corresponding incident spectrum f (lambda) by using a neural network algorithm according to the optimized measurement value.

Further objects and advantages of the present invention will become fully apparent from the following description and the accompanying drawings.

These and other objects, features and advantages of the present invention will become more fully apparent from the following detailed description and accompanying drawings.

Drawings

Fig. 1 is a schematic diagram of a spectrum chip according to a first preferred embodiment of the invention.

Fig. 2A to 2C are schematic views of the microstructure of the spectrum chip according to the above preferred embodiment of the invention.

Fig. 3A and 3B are schematic diagrams of determining a texture direction by the spectrum chip according to the above preferred embodiment of the present invention.

Fig. 4 is a schematic diagram of the spectrum chip according to the preferred embodiment of the invention for determining the direction of texture by adopting a gradual change method.

Fig. 5 is a schematic diagram of the spectrum chip for determining the texture direction by imaging to determine the texture direction according to the above preferred embodiment of the present invention.

Fig. 6A to 6C are schematic diagrams of the spectrum chip determining relevant common pixel sets according to the above preferred embodiment of the present invention.

Fig. 7A to 7B are schematic diagrams of another alternative implementation of the spectrum chip to determine relevant common pixel sets according to the above preferred embodiment of the present invention.

Fig. 8 is a schematic diagram of the method steps of a method for extracting light intensity information of a split-region spectrum chip according to the above preferred embodiment of the present invention.

Detailed Description

The following description is presented to enable one of ordinary skill in the art to make and use the invention. The preferred embodiments in the following description are by way of example only and other obvious variations will occur to those skilled in the art. The basic principles of the invention defined in the following description may be applied to other embodiments, variations, modifications, equivalents, and other technical solutions without departing from the spirit and scope of the invention.

It will be appreciated by those skilled in the art that in the present disclosure, the terms "longitudinal," "transverse," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," etc. refer to an orientation or positional relationship based on that shown in the drawings, which is merely for convenience of description and to simplify the description, and do not indicate or imply that the apparatus or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore the above terms should not be construed as limiting the present invention.

It will be understood that the terms "a" and "an" should be interpreted as referring to "at least one" or "one or more," i.e., in one embodiment, the number of elements may be one, while in another embodiment, the number of elements may be plural, and the term "a" should not be interpreted as limiting the number.

Referring to fig. 1 to 8 of the drawings, a light intensity information extraction method of a split-region spectrum chip according to a first preferred embodiment of the present application is set forth in the following description. The spectrum chip comprises a light filtering structure 10 and an image sensor 20, wherein the light filtering structure 10 is positioned on a photosensitive path of the image sensor 20, and the light filtering structure 10 is a broadband light filtering structure on a frequency domain or a wavelength domain. The transmission spectra of different wavelengths of the filter structure are not identical. In the preferred embodiment of the present application, the optical filtering structure 10 may be a structure or material having optical filtering characteristics such as a super surface, a photonic crystal, a nano-pillar, a multi-layer film, a dye, a quantum dot, a MEMS (micro electro mechanical system), an FP etalon, a cavity layer, a waveguiding layer, a diffraction element, or the like. For example, in the embodiment of the present application, the optical filtering structure 10 may be a light modulation layer in chinese patent CN 201921223201.2.

The image sensor 20 may be a CMOS Image Sensor (CIS), a CCD, an array photodetector, or the like. The spectroscopic device further includes a signal processing unit 30, and the signal processing unit 30 may be a processing unit such as MCU, CPU, GPU, FPGA, NPU, ASIC, which can export data generated by the image sensor 20 to the outside for processing.

By way of example, in this preferred embodiment of the present application, the filter structure is implemented as a light modulation layer in the following description. Specifically, the light modulation layer includes a plurality of structural units arranged in a specific pattern, wherein the structural units are arranged in a certain period. Further, the structural unit is formed by at least one micro-nano structure, wherein the micro-nano structure can be implemented as a hole, a column, a line, or the like.

The present application further provides the working principle of the spectrum chip, when the image sensor 20 measures the spectrum response, the spectrum response is transmitted into the data processing unit 30 for recovery calculation. The process is specifically described as follows:

the intensity signal of the incident light at different wavelengths λ is denoted as x (λ), the transmission spectrum of the light modulation layer is denoted as T (λ), the light modulation layer has m structural units thereon, the transmission spectrum of each structural unit is different from each other, and thus the light modulation layer is denoted as Ti (λ) (i=1, 2,3, …, m). The light intensity bi modulated by the light modulation layer is detected by a corresponding physical pixel under each structural unit. By way of example, in certain embodiments of the present application, one structural element is corresponding to one physical pixel. Alternatively, in other embodiments of the present application, a plurality of physical pixels may be a group corresponding to one structural unit. Thus, in a computing spectroscopic device according to an embodiment of the present application, at least two of the structural units constitute one "spectroscopic pixel", i.e. a plurality of structural units and corresponding image sensors constitute a spectroscopic pixel.

It will be appreciated that the number of effective transmission spectra (transmission spectra for spectral recovery, called effective transmission spectra) Ti (λ) of the light modulation layer may not be consistent with the number of structural units, and the transmission spectra of the light filtering structure 10 may be manually set, tested, or calculated according to a certain rule according to the identified or recovered requirement (for example, the transmission spectra of each structural unit obtained by testing is the effective transmission spectra). The number of effective transmission spectra of the light modulating layer may be smaller than the number of structural units, and possibly even larger than the number of structural units. Alternatively, in a specific example of the present application, a certain of the transmission spectrum curves is not necessarily determined by one structural unit.

The relationship between the spectral distribution of the incident light and the measured value of the image sensor can be expressed by the following equation:

bi＝∫x(λ)*Ti(λ)*R(λ)dλ

discretizing to obtain

bi＝Σ(x(λ)*Ti(λ)*R(λ))

Where R (λ) is the response of the image sensor 20, denoted as:

Ai(λ)＝Ti(λ)*R(λ)，

the above equation can be extended to a matrix form:

wherein bi (i=1, 2,3, …, m) is the response of the image sensor after the light to be measured passes through the light modulation layer, and corresponds to the light intensity measurement values of the image sensor corresponding to the m structural units, respectively. The matrix A (transmission spectrum curve or transmission spectrum matrix) is the light response of the system for different wavelengths, and is determined by two factors, namely the transmittance of the light modulation layer and the quantum efficiency of the image sensor. A is a matrix, each row vector corresponds to the response of a structural unit to incident light with different wavelengths, and the incident light is discretely and uniformly sampled, and n sampling points are all used. The column number of a is the same as the number of samples of the incident light. x (λ) is the intensity of the incident light at different wavelengths λ, i.e. the spectrum of the incident light to be measured.

On the basis of the implementation mode, the spectral pixels are subjected to array processing, so that the snapshot type spectral imaging device can be realized.

However, single imaging or single spectrum restoration and identification cannot be satisfied for individual scene applications, but to realize two functions simultaneously, multiple chips are generally required to work together, for example, a spectrum chip and a CMOS chip work cooperatively, which makes the volume and the cost too large. Based on this, the spectrum chip of the preferred embodiment of the present invention has a modulation region 110 and a non-modulation region 120, wherein the modulation region 110 is provided with the optical filtering structure 10, the optical filtering structure at least comprises a structural unit, wide spectrum modulation of the incident light is achieved by the optical filtering structure 10, and the non-modulation region 120 is not provided with the optical filtering structure 10. That is, the structural units of the modulation region 110 and the corresponding physical pixels constitute structural pixels, and the physical pixels of the non-modulation region 120 are called normal pixels. Taking a CMOS sensor as an example, the incident light in the modulation region 110 enters the optical filtering structure 10 to be modulated, and then enters the CMOS physical pixel corresponding to the modulation region 110 to obtain light intensity information, thereby obtaining spectrum information; the non-modulated area 120 can directly enter the corresponding physical pixel without modulating the incident light, and obtain the corresponding light intensity information, so as to obtain the image information and the like. The non-modulated regions 120 are implemented as voids, bayer arrays (regular or irregular arrays), microlenses, etc., such as voids. By way of example, the non-modulated regions 120 are implemented as black and white pixels, such as bayer arrays, the non-modulated regions are implemented as RGGB arrays, or the like.

Preferably, the modulation region 110 of the spectrum chip in the embodiment of the present application is formed in a central region of the spectrum chip, the non-modulation region 120 is formed around the spectrum chip, and the non-modulation region 120 at least partially surrounds the modulation region 110. Optionally, the modulation region 110 of the spectrum chip is formed around the spectrum chip, and the non-modulation region is formed in a central region of the spectrum chip. Optionally, in another optional embodiment of the present application, the modulation region 110 and the non-modulation region 120 of the spectrum chip are spaced apart. As an example, one structural unit corresponds to one physical pixel, and not limiting, one structural unit may correspond to a plurality of physical pixels, for example, 1 structural unit may correspond to 2×2,3×3 physical pixels or more physical pixels.

The spectrum chip modulates incident light, wherein the structural units have different transmittances for different wavebands, the modulation effects of the different structural units are different, and the spectrum information can be recovered by calculating by utilizing the different transmittances. The spectrum chip of the preferred embodiment can acquire both images and corresponding spectrum information. If the object to be identified is not necessarily regular, such as a fingerprint, a certain texture may exist, at this time, after the light source irradiates the object to be identified, incident light is generated and enters the spectrum chip, and a certain difference may exist between the incident light acquired by the structural pixel and the incident light acquired by the ordinary pixels around the structural pixel, and if the substance identification or the spectrum recovery is simply performed based on the spectrum information acquired by the structural pixel, the result may be inaccurate. If the spectrum chip needs to have a clear image and needs to acquire spectrum information, the influence of the photographed texture features on the light intensity information needs to be considered. Since there may be a large difference between ordinary pixels around the structural pixels, a simple method such as averaging cannot be considered for processing. Based on the method, the corresponding common pixels are selected through texture identification, average values are obtained based on the measured values of the common pixels, and the measured values of the corresponding structural pixels are optimized based on the average values, so that identification precision or spectrum recovery precision is improved.

As shown in fig. 3A and 3B, by way of example, in the preferred embodiment of the present application, the modulation region 110 and the non-modulation region 120 of the spectrum chip are disposed at a distance from each other. When the object to be identified has a texture (such as a fingerprint), the incident light reaches the spectrum chip and then presents different brightness conditions in the region according to the texture conditions, wherein the brightness of different regions of the texture is different, and the measured value of the corresponding position is relatively large in size difference.

As shown in fig. 3B, the intensity is more uniform along the direction of the texture, while the corresponding intensity varies, possibly gradually decreases or gradually increases, along the direction perpendicular to the texture. If the signal processing unit 30 of the spectral chip optimizes the measurement values of the structure pixels based on the measurement values of the normal pixels surrounding the structure pixels, this may make the contrast of the texture insufficient, which is obviously not reasonable.

As shown in fig. 3B, by way of example, the direction indicated by the arrow L in the drawing is the direction along the texture of a structural pixel in the modulation region, and the direction indicated by the arrow H in the drawing is the direction along which the structural pixel in the modulation region is vertically textured. It is apparent that the effect of the measurement of the normal pixels around the structural pixel along the direct texture direction is more important for the current structural pixel than for the other directions.

As shown in fig. 3A to 8, the light intensity information extraction method of the spectrum chip includes the following steps:

step S1, determining the texture direction of an object to be identified;

As shown in fig. 4, in step S1 of the light intensity information extraction method according to the present invention, the texture direction of the object to be identified is determined by a gradual change method. Specifically, a gradient algorithm is adopted to calculate the local brightness gradient of the spectrum chip, and the direction perpendicular to the gradient direction is the texture direction. The gradient direction of the texture image with the identification object is perpendicular to the texture direction.

As shown in fig. 5, in step S1 of the light intensity information extraction method according to the present invention, the texture direction of the object to be identified is determined by an imaging method. Specifically, the spectrum chip is used for imaging the identification object, a texture image is obtained, and the texture direction is identified based on the texture image. Fig. 5 shows a texture image of a fingerprint, from which the corresponding texture direction can be clearly obtained.

As shown in fig. 6A to 6C, step S2 in the light intensity information extraction method of the spectrum chip described in the present application further includes:

step S2.1, determining two sampling points (common pixels) according to the set sampling distance and the texture direction; and

In the step S2, the sampling points are normal pixels located at two sides of the structural pixel along the texture direction. As an example, in step S2.1, the side length of the physical pixel is set with the sampling distance equal to 1.5 times. As an example, in step S2.2, ten normal pixels are taken to constitute the normal pixel set; or six normal pixels may be taken to form the normal pixel set. That is, four normal pixels, which take two neighboring corresponding structure pixels, are redetermined based on the sampling points.

The step S2 in the light intensity information extraction method of the spectrum chip further includes:

and S2.3, removing common pixels with excessive difference of measured values in the common pixel set, and removing structural pixels adjacent to the sampling points. It can be understood that if the measured value of an individual normal pixel in the normal pixel set is too large, the normal pixel is removed; in addition, if the pixels adjacent to the sampling points are structural pixels, the pixels need to be removed, and the result is affected by different surfaces.

The step S3 in the light intensity information extraction method of the spectrum chip further includes:

step S3.1, processing a measured value of a common pixel set by a averaging method and obtaining a processing value ci corresponding to the structural pixel; and

step S3.2, optimizing the measured value bi of the structural pixel according to the processed value ci, and obtaining an optimized measured value bi'.

Specifically, in a specific example of the present application, in step S3 of the extraction method, the measurement value of the structural pixel is bi, and the measurement value cij of the normal pixel set is further determined according to step S2. Processing the measured value of the common pixel set in a mode of averaging the average value, the geometric average value and the like to obtain a processed value ci; and obtaining an optimized measured value bi' by optimizing the measured value bi of the structural pixels according to the processing value ci. As an example, in the preferred embodiment of the present application, bi/ci may be taken to obtain bi ', bi-ci may also be taken to obtain bi', and so on. It should be noted that the number of the normal pixel sets is determined according to step 2, wherein the number of the normal pixel sets is greater than or equal to 3.

As an example, the average value of each sampling point and the adjacent normal pixels may be calculated first, and then the average value of two sampling points may be averaged to obtain the corresponding processing value ci. The average value of the sampling point and the adjacent normal pixels can be used as the weight of the corresponding normal pixel according to the distance s from the center of the normal pixel to the end point of the sampling distance of the center of the structural pixel along the texture direction, for example, the weight is the inverse (1/s) of the distance, and then the corresponding average value is calculated.

In a specific example of the present application, the measured value is optimized, the measured value corresponding to the common pixels (sampling points) around each structural pixel is recorded as cij, for example, each sampling point is surrounded by 4 common pixels, the measured values of 8 common pixels and 2 sampling points can be averaged to obtain an average value ci, then bi/ci is obtained to obtain an optimized measured value bi 'of the structural pixel, and then the optimized measured value bi' is used for spectrum recovery. It should be noted that, in the preferred embodiment of the present application, the number of sampling points, i.e. the common pixels, is only used herein as an example, and not as a limitation. Therefore, the measurement value averaging of 10 normal pixels (including 2 sampling points) is merely for illustration, and not limitation, and the corresponding normal pixels may be determined by a specific method, and the measurement value of the structural pixels may be optimized based on the determined measurement value of the normal pixels, where the normal pixels may be greater than 10, for example, a plurality of normal pixels are spaced between the structural pixels; it is also possible to have less than 10, for example, a plurality of structural pixels are disposed adjacent to each other, and the number of adjacent or closer normal pixels is less than 10. Alternatively, the optimized measurement values of the structural pixels may not be averaged, for example, the closer the physical pixels of the structural pixels are to the higher the ratio, the lower the ratio of the distances are, and the average value ci thereof is determined. Further, the optimized measurement value bi' of the structural pixel is the measurement value bi minus the average value ci.

The light intensity information extraction method of the spectrum chip of the preferred embodiment of the invention further comprises the following steps:

step S4 obtains a corresponding incident spectrum f (λ) based on the obtained optimized measurement bi' of the structural pixel.

In the step S4 of the light intensity information extraction method of the present application, a corresponding incident spectrum f (λ) is calculated using bi' and a corresponding transmission spectrum matrix Ai (λ); or recovering the corresponding incident spectrum f (lambda) by using a neural network algorithm according to bi'.

Referring to fig. 7A and 7B of the drawings of the present specification, another alternative embodiment of a light intensity information extraction method of a spectrum chip according to the present application is set forth in the following description. The difference from the first preferred embodiment described above is in step S2 and step S3.

Specifically, step S2 of the light intensity information extraction method further includes:

As an example, in the preferred embodiment of the present application, a center point of a structural pixel and a center point of a certain common pixel around the structural pixel are selected, a distance d between the two points and an included angle θ between a line segment formed by the two points and a gradient direction are determined, and then the weight of the point is w= |sin (θ) |/d. For example, taking j common pixels around the structural pixel to form a common pixel set, the weight wj= |sin (θj) |/dj of the corresponding common pixel. It should be noted that, in the preferred embodiment of the present application, the reference luminance (processing value) of the ith structural pixel is denoted as ci= (c1×w1+c2×w2+ … +cij×wj)/(w1+ … +wj).

In order to obtain the common pixel set more accurately, the weight can be judged, so that error information is reduced. Therefore, the step S2 of the light intensity information extraction method further includes:

As shown in fig. 7B, by setting the threshold x, for example, the threshold is 0.25, if the threshold corresponding to the normal pixel is less than 0.25, the normal pixel is removed, and the normal pixel with the weight equal to or greater than 0.25 is selected and determined as the normal pixel set. As an example, if the finally determined normal pixel set is j normal pixels, the processing value ci= (c1×w1+c2×w2+ … +cij×wj)/(w1+ … +wj). In fig. 7B, the diagonal cross-hatching area represents a normal pixel having a weight of 0.25 or more, and the square cross-hatching area represents a normal pixel having a weight of less than 0.25.

As an example, in the preferred embodiment of the present application, the weights are calculated for pixel points within a distance of adjacent 3 pixel points (normal pixels or structural pixels). It should be noted that the structural pixels do not participate in the calculation when the weights are calculated.

The step S3 of the light intensity information extraction method further includes:

step S3.3, obtaining a processing value ci according to the measured value of each common pixel and the weight thereof; and

step S3.4 optimizes the measurement value bi of the structural pixels according to the processing value ci, and obtains an optimized measurement value bi'.

As an example, as shown in fig. 7A and 7B (partial structures and PD are not shown in the figures), for example, the normal pixel 1 corresponds to an angle θ1, a distance pair d1, and a weight thereof is w1= |sin (θ1) |/d1; the angle corresponding to the common pixel 2 is theta 2, and the distance pair d2 has the weight of w2= |sin (theta 2) |/d2; the angle corresponding to the common pixel 3 is theta 3, and the distance pair d3 has the weight of w3= |sin (theta 3) |/d3; the angle corresponding to the common pixel 4 is theta 4, and the distance pair d4 has the weight of w4= |sin (theta 4) |/d4.

Preferably, in this preferred embodiment of the present invention, the spectrum chip may be applied to a fingerprint recognition system, particularly a living fingerprint recognition system, wherein the fingerprint recognition system comprises a light source and a fingerprint module, and the fingerprint module further comprises the spectrum chip of the preferred embodiment of the present application. In this application, the spectrum chip includes a modulation region 110 and a non-modulation region 120, where the modulation region is provided with the filter structure 10, and the non-modulation region is not provided with the filter structure 10, taking a spectrum chip disposed at intervals as an example. The light source emits light to the finger, part of the light is absorbed and part of the light is transmitted, and part of the light is reflected after reaching the finger to form incident light, part of the incident light enters the modulation area to be modulated by the corresponding light filtering structure 10, the incident light is received by the image sensor of the spectrum chip to obtain a corresponding measured value bi, and part of the incident light directly enters the non-modulation area of the spectrum chip to be received by the image sensor to obtain a corresponding measured value cij.

It should be noted that, since the finger has a fingerprint, the reflection conditions of the valley and the ridge are different after the light source irradiates the finger, for example, the valley region may reflect, the ridge region may absorb light, and it may result in that a certain structural pixel bm corresponds to the valley region, and the peripheral normal pixels bmj may correspond to the valley region and the ridge region, respectively. The texture obtaining direction can be based on the gradual change method, the fingerprint texture image can be recovered initially, and the texture judgment can be carried out according to the texture image.

And (3) optimizing the measured value of the corresponding structural pixel by the measured value of the selected common pixel set, and identifying whether the structural pixel is a living body or not according to the optimized measured value, or recovering the spectrum curve according to the optimized measured value. In other optional embodiments of the present application, the spectrum chip may perform auxiliary imaging based on the optimized measurement value, so as to improve imaging quality; because the spectrum chip partial area has the structural pixels, the measured values of the structural pixels and the common pixels can be greatly different, and the image can be restored based on the optimized measured values, so that the influence of the structural pixels on imaging is compensated.

It will be appreciated by persons skilled in the art that the embodiments of the invention described above and shown in the drawings are by way of example only and are not limiting. The objects of the present invention have been fully and effectively achieved. The functional and structural principles of the present invention have been shown and described in the examples and embodiments of the invention may be modified or practiced without departing from the principles described.

Claims

1. The light intensity information extraction method of the split-region spectrum chip is characterized by comprising the following steps of:

step S1, determining the texture direction of an object to be identified;

2. The light intensity information extraction method according to claim 1, wherein in the step S1, a gradient algorithm is adopted to calculate a local brightness gradient of the spectrum chip, and a direction perpendicular to the gradient direction is a texture direction.

3. The light intensity information extraction method according to claim 1, wherein in the step S1, an identification object is imaged by the spectrum chip, a texture image is obtained, and the texture direction is identified based on the texture image.

4. A light intensity information extraction method according to claim 2 or 3, wherein step S2 in the light intensity information extraction method further comprises:

5. The light intensity information extraction method according to claim 4, wherein step S2 in the light intensity information extraction method further comprises:

6. The light intensity information extraction method according to claim 4, wherein step S3 in the light intensity information extraction method further comprises:

7. The light intensity information extraction method of claim 6, wherein the optimized measurement value of the structure pixel is a measurement value of the structure pixel minus a averaged processed value or a ratio of the measurement value of the structure pixel to the averaged processed value.

8. A light intensity information extraction method according to claim 2 or 3, wherein step S2 in the light intensity information extraction method further comprises:

s2.5, calculating the weight of each common pixel point according to the distance between the center point of the structural pixel and each common pixel center point and the included angle theta between the line segment formed by the two points and the gradient direction.

9. The light intensity information extraction method according to claim 8, wherein step S2 in the light intensity information extraction method further comprises:

10. The light intensity information extraction method according to claim 8, wherein step S3 of the light intensity information extraction method further comprises:

step S3.3, obtaining a processing value corresponding to the structural pixel according to the measured value of each common pixel and the weight of each common pixel; and

and step S3.4, optimizing the measured value of the structural pixel according to the processing value, and obtaining the optimized measured value of the structural pixel.

11. The light intensity information extraction method according to claim 1, wherein the light intensity information extraction method further comprises:

12. The light intensity information extraction method according to claim 11, wherein the corresponding incident spectrum f (λ) is calculated using the optimized measurement value and the corresponding transmission spectrum matrix; or recovering the corresponding incident spectrum f (lambda) by using a neural network algorithm according to the optimized measurement value.