An image sensor, an imaging module and an electronic device

Abstract

The present disclosure provides an image sensor, an imaging module and an electronic device, and belongs to the technical field of optical imaging. The image sensor comprises a plurality of arrayed photosensitive modules and a plurality of corresponding reflection units, wherein the photosensitive module comprises a photosensitive unit and a readout circuit; the reflection unit is arranged at intervals with the readout circuit, and is used for reflecting at least part of the light transmitted from the readout circuit to the photosensitive unit, so as to enhance the effective light receiving efficiency of the photosensitive unit. The filling factor can be effectively improved.

Description

Technical Field

[0001] This disclosure relates to the field of optical imaging technology, and in particular to an image sensor, imaging module and electronic device. Background Technology

[0002] Image sensors, as the core semiconductor components for photoelectric conversion, use complementary metal-oxide-semiconductor (CMOS) technology to convert light signals into electrical signals. Their performance directly affects the imaging quality of fields such as cameras, smartphones, medical imaging equipment, and autonomous driving systems. The core working principle of CMOS is to capture incident photons and generate charges through a photodiode array, then amplify the signals and perform analog-to-digital conversion through a readout circuit, ultimately generating processable digital image data.

[0003] In the evolution of sensor architecture, stacked design utilizes three-dimensional vertical integration technology to stack the photosensitive layer (pixel array) and signal processing layer (logic circuits, memory cells) at the wafer level, and uses through-silicon vias (TSVs) to achieve interlayer interconnection. This three-dimensional architecture not only significantly improves chip area utilization, but also effectively avoids the performance bottlenecks of traditional planar structures by physically isolating the photosensitive area and the circuit area, and has now become a key technology path for high-end mobile imaging and automotive vision systems.

[0004] Although traditional stacked image sensors, such as back-illuminated stacked sensors, have achieved optical path optimization, the peripheral circuits (metal interconnect layers) and non-photosensitive components such as transistors occupy 30% to 70% of the area of ​​the pixel unit, making it difficult for the fill factor (i.e. the percentage of the effective photosensitive area of ​​the photosensitive region to the total area of ​​a single pixel) to exceed the 70% threshold. However, the low fill factor leads to a decrease in photon utilization, resulting in poor image quality. Summary of the Invention

[0005] This disclosure provides an image sensor, imaging module, and electronic device that can effectively improve the fill factor.

[0006] The technical solution disclosed herein is implemented as follows: In a first aspect, this disclosure provides an image sensor comprising a plurality of photosensitive modules arranged in an array and a reflective unit corresponding to each photosensitive module. Each photosensitive module includes a photosensitive unit and a readout circuit. The reflective unit is spaced apart from the readout circuit and is used to reflect at least a portion of the light transmitted from the readout circuit to the photosensitive unit to enhance the effective light receiving efficiency of the photosensitive unit.

[0007] Secondly, this disclosure also provides an imaging module, the photosensitive module including an image processing unit and an image sensor as described in the first aspect; the image processing unit is used to generate image data based on the digital signal output by the image sensor.

[0008] Thirdly, this disclosure also provides an electronic device that includes the aforementioned imaging module.

[0009] This disclosure provides an image sensor in which a reflective unit is vertically disposed on the backlight side (Z-axis direction) of the readout circuit and maintains a preset distance from the readout circuit. By using a dual-path light capture mechanism of direct reception and reflected light recovery, the physical limitation of the fill factor is overcome, thereby increasing the effective photosensitive area of ​​the photosensitive unit. This improves the fill factor while maintaining the compatibility of traditional stacked image sensor technology, enabling the generation of high-quality images even in low-light environments. Attached Figure Description

[0010] Figure 1 This is a schematic diagram of the structure of a conventional back-illuminated image sensor provided in this disclosure.

[0011] Figure 2 This is a schematic diagram of the structure of an image sensor provided in this disclosure.

[0012] Figure 3 This is a schematic diagram of the structure of an image sensor including two concave mirrors provided in this disclosure.

[0013] Figure 4 This is a schematic diagram of the structure of an image sensor including a light propagation path, as provided in this disclosure.

[0014] Figure 5 This is a schematic diagram of another image sensor provided in this disclosure.

[0015] Figure 6 This is a schematic diagram of the principal ray angle provided in this disclosure.

[0016] Figure 7 This is a schematic diagram of the boundary and central region of the photosensitive unit provided in this disclosure.

[0017] Figure 8 This is a schematic diagram illustrating the design principle of a first concave reflector provided in this disclosure.

[0018] Figure 9 This is a schematic diagram illustrating the design principle of another first concave reflector provided in this disclosure.

[0019] Figure 10 This is a schematic diagram illustrating the design principle of a second concave reflector provided in this disclosure. Detailed Implementation

[0020] The technical solutions in the embodiments of this disclosure will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure are within the scope of protection of this disclosure.

[0021] The terms “first,” “second,” etc., used in this disclosure are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that this disclosure can be implemented in orders other than those illustrated or described herein, and the objects distinguished by “first,” “second,” etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more.

[0022] like Figure 1 As shown, a single pixel unit 10 of a conventional back-illuminated stacked image sensor consists of a photosensitive unit 101 (photodiode) and a photoelectric conversion control circuit 401 (pixel transistor) arranged side-by-side in the XY plane, while the peripheral circuit 402 is stacked vertically below the pixel array along the Z-axis. The working principle and structural characteristics of a conventional back-illuminated stacked image sensor can be broken down as follows: Located on the top layer of the semiconductor material (on the light incident side), the incident light is focused by a microlens and spectrally filtered by a color filter. The photon energy is absorbed by the silicon-based material of the photosensitive unit 101, exciting electron-hole pairs and forming a charge packet proportional to the light intensity, thus completing the physical conversion from light signal to charge signal. The photoelectric conversion control circuit 401, composed of a reset transistor, a selection transistor, and an amplification transistor, controls the charge storage and transfer timing of the photosensitive unit 101, performs charge-to-voltage conversion (through a floating diffusion node), and suppresses dark current noise and inter-pixel crosstalk. The peripheral circuit 402 is used for charge signal quantization and digitization, global exposure synchronization and readout timing management, and image data preprocessing (such as noise reduction and gain adjustment).

[0023] Combination Figure 1 It can be seen that in the traditional back-illuminated architecture, although the photosensitive unit 101 is placed at the forefront of the light incident path, it is limited by the structural constraints of the photoelectric conversion control circuit 401 and the peripheral circuit 402 integrated on the same plane, resulting in a limited effective photosensitive area ratio (fill factor). Figure 1The area of ​​the photosensitive unit 101 is only 30% to 70% of the combined area of ​​the photosensitive unit 101 and the photoelectric conversion control circuit 401. This low fill factor directly leads to a decrease in quantum efficiency and an increase in noise-equivalent illuminance. In low-light scenarios, the limited photosensitive area significantly reduces the number of photons that can be captured, causing the signal intensity to approach the background noise level of the readout circuit. Photons scattered into non-photosensitive areas may excite random charges, exacerbating fixed-mode noise and thermal noise, resulting in a sharp decline in image quality in low-light environments.

[0024] To address the aforementioned problems, this disclosure aims to provide an image sensor capable of effectively improving the fill factor. This image sensor includes multiple arrayed photosensitive modules and corresponding reflective units for each photosensitive module, such as... Figure 2 As shown, a photosensitive module 40 and a corresponding reflective unit 30 provided in this disclosure are illustrated. Each photosensitive module 40 includes a photosensitive unit 101 and a readout circuit 102. The reflective unit 30 is disposed at a distance from the readout circuit 102 and is used to reflect at least a portion of the light transmitted from the readout circuit 102 to the photosensitive unit 101 to enhance the effective light receiving efficiency of the photosensitive unit 101.

[0025] The photosensitive unit 101 is a photosensitive element (such as a photodiode) used to convert incident light into a charge signal. The readout circuit 102 is used to convert the charge signal into a digital signal. The reflection unit 30 is an optical element corresponding to the photosensitive module 40, which reflects the previously scattered light back to the photosensitive unit 101 through a specific geometric structure (such as a concave mirror or prism). Figure 2 The concave mirror is shown with the reflecting unit 30 in the middle.

[0026] The reflective unit 30 and the readout circuit 102 are spaced apart by a predetermined distance in the vertical direction (Z-axis direction shown in the figure) so that the portions that do not overlap with the projection of the photosensitive unit 101 in the vertical direction (such as...) Figure 2 The transmitted light corresponding to the incident light (the part not covered by the photosensitive unit 101 to the right of the dotted line shown in the figure) can also be reused by the photosensitive unit 101 through the reflection of the reflection unit 30.

[0027] The spacing (Z-axis distance) between the reflective unit 30 and the readout circuit 102 affects the transmission path of the reflected light, thus affecting the effective light reception efficiency. For example, a reference distance is determined based on the curvature of the reflective unit 30 and the target focal length. If the focal length of the reflective unit 30 is 5μm and the thickness of the photosensitive module 40 is 2μm, the reference distance can be set to 5μm-2μm=3μm, and then adjusted around 3μm as needed to obtain the final spacing. The specific spacing can also be determined based on the mapping relationship between the actual spacing and the effective light reception efficiency, which is not limited in this disclosure.

[0028] Effective light receiving efficiency refers to the proportion of incident light energy actually captured by the photosensitive unit 101 to the total incident light energy. With the area of ​​the photosensitive unit 101 remaining constant, the light transmitted by the readout circuit 102 is reflected back to the photosensitive unit 101 and reused by it, thus increasing the incident light energy while keeping the total incident light energy constant, thereby improving the effective light receiving efficiency. For example... Figure 2 In the incident light 1 shown, a portion of the light is transmitted through the readout circuit 102 to the reflection unit 30, and the reflection unit 30 reflects the transmitted light back to the photosensitive unit 101, so that it is captured by the photosensitive unit 101 a second time.

[0029] In some embodiments, the reflection unit 30 is configured to reflect at least a preset percentage threshold of light transmitted by the readout circuit 102 to the photosensitive unit 101. For example, if the preset percentage threshold is 70%, then at least 70% of the light energy of the transmitted light is reflected to the photosensitive unit 101.

[0030] The preset ratio threshold is used to adjust the optical design (such as concave mirror curvature optimization) or material selection (such as high reflectivity metal coating) of the reflection unit 30 to ensure that at least a preset ratio threshold (such as 70%) of the light energy transmitted from the readout circuit 102 is directionally reflected to the photosensitive unit 101, thereby significantly improving the effective light receiving efficiency of the photosensitive unit 101.

[0031] The preset ratio threshold can be determined based on the application scenario of the image sensor. If the image sensor is used in conventional camera equipment, the preset ratio threshold can be set to a smaller value, such as 50%. If the image sensor is used in night vision devices, the preset ratio threshold can be set to a larger value, such as 80%. The preset ratio threshold can also be 80%, 90%, etc., or it can be customized according to requirements.

[0032] The image sensor disclosed herein has a reflective unit 30 vertically disposed on the backlight side (Z-axis direction) of the readout circuit 102 and maintaining a preset distance from the readout circuit 102. Through a dual-path light capture mechanism of direct reception and reflected light recovery, it overcomes the physical limitations of the fill factor, thereby increasing the effective photosensitive area of ​​the photosensitive unit 101. While maintaining compatibility with traditional stacked image sensor processes, it achieves an improved fill factor, enabling the generation of high-quality images even in low-light environments.

[0033] In some embodiments, the reflective unit 30 includes at least one concave reflector, configured to be at a predetermined distance from the readout circuit 102 in a direction perpendicular to the photosensitive surface of the photosensitive unit 101. For example... Figure 3 As shown, the exemplary reflective unit 30 includes a first concave reflector 301 and a second concave reflector 302.

[0034] The special curved surface structure (such as parabola or sphere) of the concave mirror can directionally reflect and focus the transmitted light incident on its surface. This allows the light that originally passed through the readout circuit 102 to be reflected by the concave mirror and then re-converged to the effective photosensitive area of ​​the photosensitive unit 101.

[0035] The preset vertical distance between the concave mirror and the readout circuit 102 is to ensure that the reflected light spot accurately covers the photosensitive unit 101, such as... Figure 3 The vertical distance between the first concave reflector 301 and the second concave reflector 302 and the readout circuit 102 is d1. If the distance is too close, the light focusing position will deviate from the photosensitive area; if it is too far, it may cause phase interference and light energy loss due to optical path difference. Therefore, a preset distance is determined by considering the light focusing position, optical path difference, and light energy loss.

[0036] In this disclosure, since the concave mirror has the ability to converge light, more light transmitted to the reflective unit 30 can be reflected to the photosensitive unit 101, thereby improving the effective light receiving efficiency of the photosensitive unit 101.

[0037] In some embodiments, such as Figure 4 As shown, the readout circuit 102 includes a photoelectric conversion control circuit 401 and a peripheral circuit 402. The photoelectric conversion control circuit 401 and the photosensitive unit 101 are arranged side by side on the same plane, and both are attached to one side of the peripheral circuit 402. The reflection unit 30 includes a first concave reflector 301 and a second concave reflector 302. The first concave reflector 301 is used to reflect at least a portion of the light transmitted to the first concave reflector 301 to the photosensitive unit 101; the second concave reflector 302 is used to reflect at least a portion of the light transmitted to the second concave reflector 302 to the photosensitive unit 101.

[0038] If the reflecting unit 30 includes only one concave mirror, due to the inherent characteristics of the concave mirror, once the reflectivity of the transmitted light reaches a threshold, no matter how the parameters of the concave mirror are optimized, the reflected light energy cannot be further improved. Therefore, by setting two concave mirrors, the effective light receiving efficiency of the photosensitive unit 101 can be further optimized.

[0039] The first concave reflector 301 is located on the backlight surface of the peripheral circuit 402 and within the orthographic projection of the photosensitive unit 101 in the vertical direction. The second concave reflector 302 is located on the backlight surface of the peripheral circuit 402 and within the orthographic projection of the photoelectric conversion control circuit 401 in the vertical direction.

[0040] The first concave reflector 301 can reflect light incident on the photosensitive unit 101, after it has passed through the photosensitive unit 101 and the peripheral circuit 402, back to the photosensitive unit 101. Figure 4 The propagation path of light 2 in the photoelectric conversion control circuit 401; or the light ray incident from the edge of the photoelectric conversion control circuit 401, after being transmitted through the photoelectric conversion control circuit 401 and the peripheral circuit 402, is deflected to the first concave reflector 301 and reflected to the photosensitive unit 101, such as... Figure 4 The propagation path of light ray 3 in the image.

[0041] The second concave reflector 302 can reflect the light incident on the photoelectric conversion control circuit 401, after being transmitted through the photoelectric conversion control circuit 401 and the peripheral circuit 402, back to the photosensitive unit 101, such as... Figure 4 The propagation path of light ray 4 in the image.

[0042] In some embodiments, the first concave reflector 301 and the second concave reflector 302 are spaced apart from the peripheral circuit 402 in the vertical direction, and the projection of the first concave reflector 301 on the photosensitive unit 101 in the vertical direction coincides with the projection of the second concave reflector 302 on the photoelectric conversion control circuit 401 in the vertical direction.

[0043] The first concave reflector 301 and the photosensitive unit 101 are projected in the vertical direction, that is, the orthographic projection of the first concave reflector 301 and the photosensitive unit 101 are coincident. The second concave reflector 302 and the photoelectric conversion control circuit 401 are projected in the vertical direction, that is, the orthographic projection of the second concave reflector 302 and the photoelectric conversion control circuit 401 are coincident. Figure 5 As shown, the first concave reflector 301 and the second concave reflector 302 are connected, and their orthographic projections coincide exactly with the orthographic projection of the current photosensitive module 40. This arrangement can maximize the reflection of light transmitted to the first concave reflector 301 and the second concave reflector 302 to the photosensitive unit 101, and the orthographic projection of the concave reflector will not intersect with the orthographic projection of the adjacent photosensitive module 40. Therefore, it will not affect the imaging of the adjacent photosensitive module 40, that is, it will not cause crosstalk.

[0044] Because a single concave mirror is limited by the reflection angle coverage and focal shift effect, even after the reflectivity is increased to the material limit, some transmitted light still cannot be effectively recovered. The dual-mirror architecture, through a partitioned reflection mechanism, uses a first concave mirror 301 primarily to reflect the transmitted light from the photosensitive unit 101 to the peripheral circuit 402, and a second concave mirror 302 primarily to reflect the transmitted light from the photoelectric conversion control circuit 401 to the peripheral circuit 402. This improves the light energy utilization of the photosensitive unit 101 without changing the basic pixel structure.

[0045] In some embodiments, the curvature and position of the first concave reflector 301 are configured to reflect the transmitted light corresponding to the negative principal ray incident light at the boundary position of the photosensitive unit 101 away from the boundary position of the photoelectric conversion control circuit 401, and the transmitted light corresponding to the positive principal ray incident light at the boundary position of the photoelectric conversion control circuit 401, to the central region of the photosensitive unit 101. The positive principal ray angle is the critical angle at which the incident light deflects clockwise relative to the vertical direction, and the negative principal ray angle is the critical angle at which the incident light deflects counterclockwise relative to the vertical direction; the value of the critical angle ranges from 0 degrees to 90 degrees.

[0046] Since not all incident light at any angle can be captured by the photosensitive unit 101, the image sensor is designed with a preset maximum allowable angle between the light ray incident from the lens onto the photosensitive surface of the photosensitive unit 101 and the normal to the photosensitive surface. This principal ray angle is the critical angle at which incident light can be captured by the photosensitive unit 101. Light rays incident at angles exceeding the principal ray angle are less likely to be captured by the photosensitive unit 101. For example... Figure 6 As shown, the normal to the photosensitive surface of the photosensitive unit 101 is the Z-axis, and θ1 is the designed principal ray angle. Therefore, light rays incident at an angle greater than θ1 (θ2) cannot be effectively utilized by the photosensitive unit 101.

[0047] The central region of the photosensitive unit 101 refers to a preset region with the geometric center of the photosensitive surface of the photosensitive unit 101 as the origin. The boundary positions of the photosensitive unit 101 refer to the four boundaries. For example... Figure 7 As shown, the thickened line around the photosensitive unit 101 indicates the boundary of the photosensitive unit 101, and the area shown in the dashed box is the central area.

[0048] like Figure 8 As shown, θ1 is the principal ray angle, boundary 1 is the boundary of the photosensitive unit 101 away from the photoelectric conversion control circuit 401, and boundary 2 is the boundary of the photosensitive unit 101 close to the photoelectric conversion control circuit 401. Ray 5 is incident on the photosensitive unit 101 from boundary 1 at θ1, and the incident angle of ray 5 is the negative principal ray angle. Ray 6 is incident on the photosensitive unit 101 from boundary 2 at θ1, and the incident angle of ray 6 is the positive principal ray angle. Part of the two rays transmitted to the first concave reflector 301 are reflected to the central region P of the photosensitive unit 101. This design ensures that any light rays incident at an angle less than or equal to the principal ray angle between boundary 1 and the central region O, when transmitted to the first concave reflector 301, are reflected back to the photosensitive unit 101. Similarly, any light rays incident at an angle less than or equal to the principal ray angle between boundary 2 and the central region O, when transmitted to the first concave reflector 301, are reflected back to the photosensitive unit 101, such as ray 7. This ensures that as much light transmitted to the first concave reflector 301 as possible is reflected back to the photosensitive unit 101.

[0049] The curvature and position of the first concave reflector 301 can be determined mathematically based on the refractive index of the photosensitive module 40, the wavelength of the incident light, the thickness of the photosensitive module 40 (the distance between the upper photosensitive surface of the photosensitive unit 101 and the peripheral circuit 402 in the Z-axis direction), and the distance between two opposite boundary positions of the photosensitive unit 101. Alternatively, it can be adjusted gradually according to a preset adjustment strategy until the transmitted light corresponding to the principal ray angle incident light at the boundary position of the photosensitive unit 101 is reflected to the central region of the photosensitive unit 101.

[0050] For example, based on the size of the photosensitive unit 101, the principal ray angle, and the refractive index of the material, the initial curvature and initial position of the first concave reflector 301 are initially determined using geometric optics principles. Using optical simulation software, with the incident light incident at the principal ray angle and incident from the boundary point (the center point of the four boundaries of the photosensitive unit 101), it is detected whether the reflected light spot is located in the central region of the photosensitive unit 101 (e.g., the central 20% area). The percentage of overlap between the reflected light spot and the central region of the photosensitive unit 101 is calculated. Coarse adjustment optimization (coordinated adjustment of curvature and position): Based on the initial curvature, the curvature is adjusted by a first preset step size to detect whether the reflected light spot shrinks or spreads towards the central region, and to screen curvature directions that improve the overlap of the light spots. The optimized curvature is fixed, and the position is adjusted by a second preset step size to determine the optimal position for light spot focusing (minimum light spot size and highest center overlap). The synergistic effect of curvature and vertical distance is detected. For example, while increasing the curvature by 5%, the position is adjusted, and the vertical distance is reduced by 10%, eliminating combinations where the light spot is scattered or deviates from the center, and retaining the optimal M (integer greater than 1) sets of parameters for fine adjustment. Fine adjustment optimization: Near the optimal curvature of the coarse adjustment, the curvature is finely adjusted by a third preset step size smaller than the first preset step size to verify whether the overlap between the light spot and the central region is greater than or equal to a first threshold. Near the optimal position of the coarse adjustment, the position is adjusted by a fourth preset step size smaller than the second preset step size until the light spot is located in the central region. Thus, by alternating the curvature and position, the light incident at the principal ray angle at the boundary position of the photosensitive unit 101 is reflected to the central region by the light transmitted to the first concave reflector 301.

[0051] In some embodiments, the first concave reflector 301 is positioned outside the reflected light path of the second concave reflector 302. This design is to prevent light reflected from the second concave reflector 302 to the photosensitive unit 101 from being blocked by the first concave reflector 301, thus preventing the reflected light from being captured by the photosensitive unit 101. Figure 9 As shown, the first concave reflector 301 is placed within the reflected light path of the second concave reflector 302, causing the light 8 to be unable to be captured by the photosensitive unit 101.

[0052] In some embodiments, the first concave reflector 301 and the second concave reflector 302 are connected. Since light energy attenuates during propagation, a shorter optical path preserves more energy. The first concave reflector 301 must be located outside the reflected light path of the second concave reflector 302. After the position of the second concave reflector 302 is determined, the first concave reflector 301 and the second concave reflector 302 are connected, resulting in the shortest light path passing through the second concave reflector 302. Figure 9 The first concave mirror 301 and the second concave mirror 302 shown in the figure are not connected. Figure 8 The first concave mirror 301 shown is connected to the second concave mirror 302.

[0053] In some embodiments, the curvature, tilt angle, and position of the second concave reflector 302 are configured to reflect the transmitted light corresponding to the incident light at the positive principal ray angle at the boundary position of the photoelectric conversion control circuit 401 away from the photosensitive unit 101 to the boundary region of the photosensitive unit 101 away from the photoelectric conversion control circuit 401, and the position of the second concave reflector 302 is configured to be closest to the vertical distance of the photoelectric conversion control circuit 401, and the positive principal ray angle is the critical angle at which the incident light deflects clockwise relative to the vertical direction.

[0054] The boundary region of the photosensitive unit 101 refers to a pre-defined area centered on the boundary. Since light energy attenuates during propagation, the shorter the optical path, the more energy is retained. Therefore, to ensure that the transmitted light corresponding to the incident light at the positive principal ray angle is reflected to the boundary region of the photosensitive unit 101, which is far from the photoelectric conversion control circuit 401, the second concave reflector 302 should be positioned as close as possible to the vertical distance from the photoelectric conversion control circuit 401.

[0055] like Figure 10 As shown, the boundary position of the photoelectric conversion control circuit 401 on the side away from the photosensitive unit 101 is boundary 3, the positive principal ray angle is the incident angle corresponding to ray 9, and the tilt angle of the second concave reflector 302 is α. A portion of ray 9 transmitted to the second concave reflector 302 is reflected and captured by the boundary region of the photosensitive unit 101. This design ensures that light rays incident on the photoelectric conversion control circuit 401 at an angle less than or equal to the principal ray angle are reflected by the second concave reflector 302 and captured by the photosensitive unit 101, such as ray 10. This ensures that as much light transmitted to the second concave reflector 302 as possible is reflected to the photosensitive unit 101.

[0056] The curvature, position, and tilt angle of the second concave reflector 302 can be determined mathematically based on the refractive index of the photosensitive module 40, the wavelength of the incident light, the thickness of the photosensitive module 40 (the distance between the upper photosensitive surface of the photosensitive unit 101 and the peripheral circuit 402 in the Z-axis direction), and the distance between two opposite boundary positions of the photoelectric conversion control circuit 401. Alternatively, it can be adjusted step by step according to a preset adjustment strategy until the transmitted light corresponding to the positive principal ray angle incident light at the boundary position of the photoelectric conversion control circuit 401 is reflected to the boundary region of the photosensitive unit 101 away from the photoelectric conversion control circuit 401. The specific process of iteratively determining the curvature, position, and tilt angle of the second concave reflector 302 can refer to the iterative process of the first concave reflector 301 described above, and will not be repeated here.

[0057] It should be noted that, since the placement of the first concave reflector 301 is related to the reflected light path of the second concave reflector 302, it is necessary to iteratively determine the position, curvature, and tilt angle of the second concave reflector 302 to ensure that it reflects light to the photosensitive unit 101, has the shortest vertical distance to the photoelectric conversion control circuit 401, and does not interfere with it. After determining the position, curvature, and tilt angle of the second concave reflector 302, since the first concave reflector 301 requires the shortest possible optical path, it is connected to the second concave reflector 302, i.e., the position of the first concave reflector 301 is determined. Then, the curvature of the first concave reflector 301 is iteratively changed to ensure that it reflects light to the photosensitive unit 101 without interference.

[0058] In some embodiments, the difference between the optical path of the light reflected to the photosensitive unit 101 and the optical path of the corresponding incident light is an integer multiple of the wavelength of the incident light.

[0059] To avoid interference between the reflected light and the incident light, which would reduce light energy, the parameters of the reflecting unit 30 must be configured such that the difference between the optical path length of the light reflected to the photosensitive unit 101 and the corresponding optical path length of the incident light is an integer multiple of the wavelength of the incident light. This ensures that there is no interference and prevents light energy loss. If the reflecting unit 30 includes a first concave reflector 301 and a second concave reflector 302, then the curvature and position of the first concave reflector 301, and the curvature, position, and tilt angle of the second concave reflector 302, must ensure that the optical path difference is an integer multiple of the incident light.

[0060] In some embodiments, the curvature and position of the first concave reflector 301 are set such that, within a first wavelength range, the actual focal point of the incident light at each wavelength is offset from the theoretical focal point of the center wavelength within the first wavelength range by an amount less than or equal to an offset threshold. Similarly, the curvature, position, and tilt angle of the second concave reflector 302 are set such that, within a second wavelength range, the actual focal point of the incident light at each wavelength is offset from the theoretical focal point of the center wavelength within the second wavelength range by an amount less than or equal to an offset threshold. The first and second wavelength ranges may not overlap, or they may partially overlap, such as the first wavelength range being the visible light band (400nm-700nm) and the second wavelength range being the near-infrared band (700nm-1700nm). The actual focal point refers to the point where reflected light is actually focused, while the theoretical focal point refers to the geometric point where a concave mirror (or lens) converges parallel incident light rays under an ideal optical model. Its position is determined by the geometry and curvature of the mirror and serves as the reference point for optical system design. The center wavelength refers to the wavelength at the midpoint of the range. For example, if the first wavelength range is 400nm-700nm, then the center wavelength is 550nm.

[0061] When light passes through the photosensitive module 40 (such as silicon dioxide or silicon nitride), the refractive index changes with wavelength. Short-wavelength light (such as 300nm ultraviolet) has a higher refractive index, while long-wavelength light (such as 1700nm infrared) has a lower refractive index, causing the focal position of light in different wavelength bands to shift at the same curvature. As the curvature increases, the focal length shortens, the light-gathering ability increases, and the depth of focus (the longitudinal range in which light remains clearly focused near the focal point) becomes shallower.

[0062] Therefore, in order to ensure that the reflected light corresponding to incident light of different wavelengths can be captured by the photosensitive unit 101, the curvature and position of the first concave mirror 301 are set such that, within the first wavelength range, the offset of the actual focal point of each wavelength from the theoretical focal point of the center wavelength within the first wavelength range is less than or equal to an offset threshold. Specifically, taking the theoretical focal point of the center wavelength (e.g., 550nm) of the first wavelength range as a reference, it is ensured that the actual focal point of any wavelength within this band satisfies the following: the lateral offset (XY plane) is less than the lateral offset threshold (e.g., 10% of the side length of the photosensitive unit 101), and the longitudinal offset (Z-axis) is less than the longitudinal offset (e.g., the thickness of the photosensitive unit 101). Limiting the offset of the actual focal point of different wavelengths from the theoretical focal point of the center wavelength in the lateral direction is to ensure that the reflected light of different wavelengths within the first wavelength range can be reflected to the first concave mirror 301. Limiting the offset of the actual focal point of different wavelengths from the theoretical focal point of the center wavelength in the longitudinal direction is to ensure that the focal point of the reflected light of different wavelengths within the first wavelength range is located within the photosensitive unit 101 in the longitudinal direction, ensuring that the reflected light can be clearly focused. Within the second wavelength range, the actual focal point of the second concave mirror 302 is offset from the theoretical focal point of the center wavelength within the second wavelength range by an amount less than or equal to the offset threshold. The principle is the same as that of the first concave mirror 301, and will not be repeated here.

[0063] In some embodiments, the curvature and position of the first concave mirror 301, and the curvature, position, and tilt angle of the second concave mirror 302, are adjusted according to the nonlinear mapping relationship between the curvature and position of the first concave mirror 301 and the curvature, position, and tilt angle of the second concave mirror 302 and the effective light receiving efficiency, until the effective light receiving efficiency of the photosensitive unit 101 meets the desired threshold.

[0064] In some implementations, a neural network model is used to determine the curvature and position of the first concave mirror 301, and the curvature, position, and tilt angle of the second concave mirror 302, corresponding to the effective light receiving efficiency. The neural network model is trained to learn the aforementioned nonlinear mapping relationship. Specifically, a training dataset is generated. Within a reasonable range (achievable by the optical structure), the curvature and three-dimensional coordinates of the first concave mirror 301, and the curvature, three-dimensional coordinates, and tilt angle of the second concave mirror 302 are randomly generated. For each set of parameters, the corresponding effective light receiving efficiency is calculated using ray tracing simulation software. Random sampling is used to obtain the training set. Each training data point includes: input data: effective light receiving efficiency (values ​​from 0% to 100%), and output data: curvature and three-dimensional coordinates of the first concave mirror 301, and curvature, three-dimensional coordinates, and tilt angle of the second concave mirror 302. The model's neural network structure (inverse prediction model) is constructed as follows: Input layer: 1 node (normalized target effective light reception efficiency value), hidden layer, multiple fully connected layers (512-1024 neurons per layer), residual connections (ResNet structure) are added to prevent gradient vanishing, activation function: ReLU is used in the first few layers, and Tanh is used in the last layer to constrain the output range, output layer: 5 nodes (corresponding to 5 optical parameters to be predicted), post-processing layer, physical constraints are applied to the output parameters (such as curvature cannot be negative, tilt angle does not exceed ±45 degrees, etc.). The training model uses a loss function including the main loss, the mean squared error between the predicted and actual parameters, and an auxiliary loss. The effective light reception efficiency error of the predicted parameters is verified using a fast forward model. The training phase consists of two stages: a coarse training phase using 80% of the data (batch size 256, learning rate 1e-3, approximately 500 epochs); and a fine training phase using the full data (batch size 64, learning rate 1e-5, approximately 200 epochs). An early stopping mechanism is included: the model terminates if the loss on the validation set does not decrease after 10 consecutive epochs. The model meets the following criteria: the average error of the predicted parameters on the validation set is less than a preset threshold (e.g., curvature error less than 0.5%, position error less than 0.1 micrometers); and the error between the predicted optical parameters (5 parameters) and the target value after simulation is less than the error threshold. Using the trained model, the desired effective light reception efficiency (e.g., inputting 0.85 represents 85%) is input, and the output is five parameter values ​​that can be directly used for optical system adjustment.

[0065] In other implementations, the curvature and position of the first concave mirror 301, as well as the curvature, position, and tilt angle of the second concave mirror 302, corresponding to the effective light receiving efficiency are determined iteratively. After determining the final desired effective light receiving efficiency, the adjustment priority of the curvature and position of the first concave mirror 301 and the curvature, position, and tilt angle of the second concave mirror 302 is determined. For example, if it is necessary to determine the parameters of the second concave mirror 302 first, and after determining the position of the second concave mirror 302, the first concave mirror 301 is connected to it, so the position of the first concave mirror 301 is also determined. Finally, the curvature of the first concave mirror 301 is determined. The specific adjustment process is illustrated using the parameter adjustment of the second concave reflector 302 as an example. First, a rapid scan is performed with a larger step size. In the optimal region, a finer step size is used for local optimization. When the adjustment is effective for m consecutive times (an integer greater than 1), the step size is automatically increased by a preset value. After the first adjustment fails, a bisection method is used for reverse search instead of simple reverse stepping. At the same time, three sets of parameter states (current optimal / forward exploration / reverse exploration) are maintained. After every 5 iterations, the optimal branch is selected to continue. After each parameter optimization stage, the full parameter state is automatically saved. Finally, the historical optimal solution is selected instead of the last iteration result until the effective receiving efficiency reaches the expected value.

[0066] In other feasible methods, since there is an optimal solution in the nonlinear mapping relationship for light within a bandwidth after the wavelength range is determined, the optimal solution can be found through the above iterative method. That is, when the parameters of the curvature, position, curvature, position, and tilt angle of the first concave mirror 301 and the second concave mirror 302 are fixed in a certain combination, the effective light receiving efficiency reaches the maximum, and this maximum value is found by gradually adjusting.

[0067] The image sensor disclosed herein recovers light that has diffused into the non-photosensitive area through a reflective unit 30. This dual-path mechanism of direct reception and reflection recovery allows the light capture range of the photosensitive unit 101 to overcome physical area limitations. The first concave reflector 301, based on the center wavelength (e.g., 550nm), constrains lateral offset (XY plane) and longitudinal depth of focus (Z-axis), ensuring that 400-700nm light is focused within the effective area of ​​the photosensitive unit 101. The second concave reflector 302 compensates for lateral dispersion through tilt angle and extends the depth of focus with a low curvature design, covering the longitudinal offset of 700-1700nm light. The optical path difference between the reflected and incident light is an integer multiple of the wavelength, suppressing energy loss due to interference and ensuring the integrity of multi-band signals. The reflective unit 30 is strictly aligned with the orthographic projection of the photosensitive unit 101 or the photoelectric conversion control circuit 401, eliminating optical path crosstalk between adjacent pixels. The reflective unit 30 is integrated into the backlight side of the readout circuit 102, compatible with traditional back-illuminated sensor processes, requiring no modification to the wafer-level packaging structure. The first concave mirror 301 captures the incident light at the principal ray angle (the critical angle between positive and negative directions) at the boundary of the photosensitive unit 101 and reflects it to the central region. The second concave mirror 302 directionally recovers the transmitted light from the photoelectric control circuit area and reflects it to the edge region of the photosensitive unit 101, expanding the effective photosensitive range. By controlling the longitudinal offset threshold, it is ensured that different wavelength focal points cover the full thickness of the photosensitive layer, reducing light energy dissipation under low illumination.

[0068] This disclosure also provides an imaging module, which includes an image processing unit and an image sensor as shown in the above embodiments; the image processing unit is used to generate image data based on the digital signal output by the image sensor.

[0069] This disclosure also provides an electronic device that includes the above-described imaging module.

[0070] It should be understood that the photosensitive module and electronic device provided in this disclosure can implement the various processes of the above-described image sensor embodiments and achieve the same technical effects. To avoid repetition, they will not be described again here.

[0071] It should be noted that the technical solutions described in this disclosure can be combined arbitrarily as long as they do not conflict.

[0072] The above description is merely a specific embodiment of this disclosure, but the scope of protection of this disclosure is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this disclosure should be included within the scope of protection of this disclosure.

Claims

1. An image sensor, characterized in that, The image sensor includes multiple arrayed photosensitive modules and a reflective unit corresponding to each photosensitive module. Each photosensitive module includes a photosensitive unit and a readout circuit. The readout circuit includes a photoelectric conversion control circuit and peripheral circuits. The photoelectric conversion control circuit and the photosensitive unit are arranged side by side on the same plane and are both attached to one side of the peripheral circuits. The reflective unit is spaced apart from the readout circuit and is used to reflect at least part of the light transmitted from the readout circuit to the photosensitive unit, thereby enhancing the effective light receiving efficiency of the photosensitive unit. The reflecting unit includes at least one concave reflector, which is positioned at a predetermined distance from the readout circuit in a direction perpendicular to the photosensitive surface of the photosensitive unit, and located on the backlight surface of the peripheral circuit. The reflecting unit includes a first concave reflector and a second concave reflector. The first concave reflector is located within the orthographic projection of the photosensitive unit in the vertical direction, and the second concave reflector is located within the orthographic projection of the photoelectric conversion control circuit in the vertical direction. The first concave reflector and the second concave reflector are spaced apart from the peripheral circuit in the vertical direction, and the projection of the first concave reflector on the photosensitive unit in the vertical direction coincides with the projection of the photoelectric conversion control circuit in the vertical direction.

2. The image sensor according to claim 1, characterized in that, The reflective unit is configured to reflect light transmitted by the readout circuit up to a percentage threshold of not less than a preset ratio to the photosensitive unit.

3. The image sensor according to claim 1, characterized in that, The curvature and position of the first concave mirror are set as follows: The transmitted light corresponding to the negative principal ray incident light at the boundary position of the photosensitive unit away from the boundary position of the photoelectric conversion control circuit, and the transmitted light corresponding to the positive principal ray incident light at the boundary position of the photoelectric conversion control circuit, are reflected to the central region of the photosensitive unit. Wherein, the positive principal ray angle is the critical angle at which the incident light deflects clockwise relative to the vertical direction, and the negative principal ray angle is the critical angle at which the incident light deflects counterclockwise relative to the vertical direction.

4. The image sensor according to claim 3, characterized in that, The first concave reflector is positioned outside the reflected light path of the second concave reflector.

5. The image sensor according to claim 4, characterized in that, The first concave reflector is connected to the second concave reflector.

6. The image sensor according to claim 1, characterized in that, The curvature, tilt angle, and position of the second concave mirror are set as follows: The transmitted light corresponding to the incident light at the positive principal ray angle at the boundary position on the side of the photoelectric conversion control circuit away from the photosensitive unit is reflected to the boundary region of the photosensitive unit away from the photoelectric conversion control circuit; and The position of the second concave reflector is set to be as close as possible to the vertical distance of the photoelectric conversion control circuit, and the positive principal ray angle is the critical angle at which the incident light deflects clockwise relative to the vertical direction.

7. The image sensor according to claim 1, characterized in that, The difference between the optical path length of the light reflected to the photosensitive unit and the optical path length of the corresponding incident light is an integer multiple of the wavelength of the incident light.

8. The image sensor according to any one of claims 1 to 7, characterized in that, The curvature and position of the first concave mirror are set as follows: Within the first wavelength range, the actual focal point of the reflected light at each wavelength is offset from the theoretical focal point of the center wavelength within the first wavelength range by an amount less than or equal to the offset threshold. The curvature, position, and tilt angle of the second concave mirror are set as follows: Within the second wavelength range, the actual focal point of each wavelength is offset from the theoretical focal point of the center wavelength within the second wavelength range by an amount less than or equal to the offset threshold.

9. The image sensor according to any one of claims 1 to 7, characterized in that, Based on the nonlinear mapping relationship between the curvature and position of the first concave mirror, and the curvature, position, and tilt angle of the second concave mirror, and the effective light receiving efficiency, the curvature and position of the first concave mirror, and the curvature, position, and tilt angle of the second concave mirror are adjusted until the effective light receiving efficiency of the photosensitive unit meets the desired threshold.

10. An imaging module, characterized in that, The photosensitive module includes: an image processing unit and an image sensor according to any one of claims 1 to 9; The image processing unit is used to generate image data based on the digital signal output by the image sensor.

11. An electronic device, characterized in that, The electronic device includes the imaging module of claim 10.

Images