Non-structural multi-modal arrayed ultra-high definition intelligent photoelectric detector
By employing a circular array of four visible light lenses and four infrared thermal imaging lenses, along with a video image stitching and fusion module, the limited field of view of traditional cameras is solved, enabling panoramic monitoring and all-weather high-definition imaging.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Utility models(China)
- Current Assignee / Owner
- SHENZHEN ZHUOHE TECH CO LTD
- Filing Date
- 2025-07-28
- Publication Date
- 2026-07-07
AI Technical Summary
Traditional monocular or binocular cameras have limited field of view, making it impossible to achieve panoramic monitoring. This results in blind spots in large-scale scenes, making it difficult to cover the entire area.
It employs a circular array consisting of four visible light lenses and four infrared thermal imaging lenses, with adjacent lenses at a 45° angle, covering a 180° horizontal field of view. Combined with a video image stitching and fusion module, it forms a panoramic video of visible light and infrared, eliminating blind spots in monitoring.
It enables panoramic monitoring of a wide range of scenes, eliminates blind spots, and provides high-definition imaging effects around the clock.
Smart Images
Figure CN224473375U_ABST
Abstract
Description
Technical Field
[0001] This utility model relates to the field of video surveillance technology, specifically to an unstructured multimodal array-type ultra-high-definition intelligent photodetector. Background Technology
[0002] Cameras installed on mobile platforms are typically used in vehicles and on ships. Vehicle-mounted and ship-mounted cameras play a crucial role in modern transportation, serving purposes such as monitoring, security, and improving driving safety. These cameras utilize high-definition lenses and are equipped with necessary image processing units to support clear imaging at night or in adverse weather conditions. They are usually equipped with various functions such as autofocus, wide-angle shooting, and night vision to adapt to different application scenarios. Furthermore, many cameras integrate intelligent analysis functions to help improve the safety of vehicle and ship operations and reduce accidents.
[0003] In the traditional field of security monitoring, monocular or binocular cameras, due to their limited field of view, can often only monitor targets within a specific range and cannot achieve panoramic monitoring. This limits their application in scenarios that require large-scale monitoring. For example, in city squares, large shopping malls, industrial parks, and other places, the shortcomings of monocular or binocular cameras become apparent, as they cannot effectively cover the entire area and it is difficult to detect and track targets in a timely manner.
[0004] In view of this, the present invention proposes an unstructured multimodal array-type ultra-high-definition intelligent photodetector to solve the above-mentioned technical problems. Utility Model Content
[0005] To address the shortcomings of the aforementioned background technologies, this utility model provides a technical solution for an unstructured multimodal array-type ultra-high-definition intelligent photoelectric detector. It employs a circular array composed of four visible light lenses and four infrared thermal imaging lenses, with adjacent lenses at a 45° angle, covering a horizontal field of view of 180° (30° for visible light and 40° for infrared). Through multi-view collaboration, it solves the problem of limited field of view in traditional monocular or binocular cameras. Combined with a video image stitching and fusion module, it seamlessly stitches together multi-channel images to form a complete visible light and infrared panoramic video, comprehensively covering large-scale scenes such as city squares and industrial parks, eliminating monitoring blind spots.
[0006] This utility model provides the following technical solution: a non-structured multi-modal array ultra-high-definition intelligent photodetector, including a multi-eye ring lens group, a dual-spectrum imaging unit, a video image stitching and fusion module, a video intelligent analysis module, a network communication module, a power management module, and a panoramic shell assembly;
[0007] The multi-lens ring array includes multiple visible light lenses and multiple infrared thermal imaging lenses, which are arranged in a ring array, and the optical axes of the visible light lenses and infrared thermal imaging lenses are radially distributed.
[0008] The dual-spectral imaging unit includes multiple visible light sensors and multiple infrared sensors. Each visible light lens is connected to a visible light sensor, and each infrared thermal imaging lens is connected to an infrared sensor, forming an independent multi-channel dual-spectral acquisition module.
[0009] The panoramic housing assembly includes a housing base, a detachable top cover, and a lens mounting bracket. The multi-lens ring lens group is embedded and fixed on the lens mounting bracket of the housing base. The visible light lens above the lens mounting bracket and the infrared thermal imaging lens below the lens mounting bracket are located on the same plane.
[0010] As a preferred embodiment of this utility model, the number of visible light lenses and infrared thermal imaging lenses is four each, with the angle between adjacent visible light lenses being 45°, covering a horizontal viewing angle of 180° and a vertical viewing angle of 30°; the angle between adjacent infrared thermal imaging lenses is also 45°, covering a horizontal viewing angle of 180° and a vertical viewing angle of 40°.
[0011] As a preferred technical solution of this utility model, the lens fixing bracket is a machined aluminum alloy bracket, which is used to ensure the horizontal angular accuracy of adjacent lenses and the stability of the optical axis, and can conduct heat to the base of the housing.
[0012] As a preferred technical solution of this utility model, the housing base and the detachable top cover of the panoramic housing assembly are respectively an aluminum alloy housing base and an aluminum alloy detachable top cover.
[0013] As a preferred embodiment of this utility model, the visible light sensor is an OS08A20 RGB imaging sensor, and the infrared sensor is an RTDS121C thermal imaging sensor.
[0014] As a preferred embodiment of this utility model, the video image stitching and fusion module is an edge fusion controller of model SOM3559A-A16C0D-A2-C, the video intelligent analysis module is a deep learning inference engine of model HT108CB100-BTB, the network communication module is a gigabit Ethernet communication module of model ZH_DZ_SW_YT9218NH, and the power management module is a power module of model KUB4812QB-10A.
[0015] As a preferred technical solution of this utility model, the multi-eye ring lens group, dual-spectrum imaging unit, video image stitching and fusion module, video intelligent analysis module, network communication module and power management module are all installed in the inner cavity of the housing base through sealing rings and anti-loosening screws.
[0016] Compared with the prior art, the present invention has the following beneficial effects:
[0017] 1. This utility model uses a circular array composed of four visible light lenses and four infrared thermal imaging lenses, with adjacent lenses at a 45° angle, covering a horizontal field of view of 180° (30° for visible light vertical field of view and 40° for infrared vertical field of view). By using multi-lens collaboration, it solves the problem of limited field of view of traditional monocular or binocular cameras. Combined with a video image stitching and fusion module, it seamlessly stitches together multi-channel images to form a complete visible light and infrared panoramic video, which can fully cover large-scale scenes such as city squares and industrial parks, and eliminate monitoring blind spots. Attached Figure Description
[0018] Figure 1 This is a schematic diagram of the structure of this utility model;
[0019] Figure 2 This is a schematic diagram of the dual-spectral lens assembly structure of this utility model;
[0020] Figure 3 This is a schematic diagram of the panoramic shell assembly structure of this utility model;
[0021] Figure 4 This is a front view of the lens fixing bracket of this utility model;
[0022] Figure 5 This is a rear view of the lens mounting bracket of this utility model.
[0023] In the diagram: 1. Visible light lens; 101. Infrared thermal imaging lens; 2. Visible light sensor; 201. Infrared sensor; 3. Housing base; 301. Detachable top cover; 302. Lens mounting bracket; 4. Video image stitching and fusion module; 5. Video intelligent analysis module; 6. Network communication module; 7. Power management module. Detailed Implementation
[0024] The technical solutions of the present utility model will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present utility model, and not all embodiments. Based on the embodiments of the present utility model, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of the present utility model.
[0025] Please see Figure 1-5 As shown, the unstructured multimodal array ultra-high-definition intelligent photodetector includes a multi-eye ring lens group, a dual-spectrum imaging unit, a video image stitching and fusion module 4, a video intelligent analysis module 5, a network communication module 6, a power management module 7, and a panoramic housing assembly.
[0026] The multi-lens ring array includes multiple visible light lenses 1 and multiple infrared thermal imaging lenses 101. The visible light lenses 1 and the infrared thermal imaging lenses 101 are arranged in a ring array, and the optical axes of the visible light lenses 1 and the infrared thermal imaging lenses 101 are distributed radially.
[0027] The dual-spectrum imaging unit includes multiple visible light sensors 2 and multiple infrared sensors 201. Each visible light lens 1 is connected to a visible light sensor 2, and each infrared thermal imaging lens 101 is connected to an infrared sensor 201, forming an independent multi-channel dual-spectrum acquisition module.
[0028] The panoramic housing assembly includes a housing base 3, a detachable top cover 301, and a lens mounting bracket 302. The multi-lens ring lens group is embedded and fixed on the lens mounting bracket 302 of the housing base 3. The visible light lens 1 above the lens mounting bracket 302 and the infrared thermal imaging lens 101 below it are located on the same plane.
[0029] The lens mounting bracket 302 is machined to ensure a high-precision horizontal positional relationship between adjacent lenses, guaranteeing a fixed angle. Simultaneously, it ensures that the visible light lens 1 above and the infrared thermal imaging lens 101 below are mounted on the same plane in the vertical direction, making their positional relationship stable and accurate, ensuring a good stitching effect. Furthermore, by sharing the lens mounting bracket 302, the heat dissipation area is increased. Additionally, the lens mounting bracket 302 is tightly connected to the housing base 3, allowing heat from the infrared sensor 202 to be conducted to the bottom 3 of the housing for heat dissipation, reducing the thermal noise of the infrared sensor 202.
[0030] There are four visible light lenses 1 and four infrared thermal imaging lenses 101. The angle between adjacent visible light lenses 1 is 45°, covering a horizontal field of view of 180° and a vertical field of view of 30°. The angle between adjacent infrared thermal imaging lenses 101 is 45°, covering a horizontal field of view of 180° and a vertical field of view of 40°.
[0031] The lens mounting bracket 302 is a machined aluminum alloy bracket used to ensure the horizontal angular accuracy of adjacent lenses and the stability of the optical axis, and to conduct heat to the housing base 3;
[0032] The panoramic housing assembly's housing base 3 and detachable top cover 301 are respectively an aluminum alloy housing base and an aluminum alloy detachable top cover.
[0033] Visible light sensor 2 is an RGB imaging sensor with model number OS08A20, and infrared sensor 201 is a thermal imaging sensor with model number RTDS121C.
[0034] The visible light sensor 2 (model: OS08A20) is manufactured by OmniVision Integrated Circuits Group, Inc. The OS08A20 is a 1 / 1.8-inch RGB imaging sensor launched by OmniVision Technology, which supports high dynamic range (HDR) and is widely used in security cameras.
[0035] The infrared sensor 201 (model: RTDS121C) is manufactured by Yantai IRay Optoelectronics Technology Co., Ltd. The RTDS121C is an infrared thermal imaging sensor 201 from Yantai IRay Optoelectronics Technology Co., Ltd. It supports 1280×1024 pixel infrared imaging and is suitable for thermal imaging security monitoring scenarios.
[0036] The video image stitching and fusion module 4 is an edge fusion controller of model SOM3559A-A16C0D-A2-C, the video intelligent analysis module 5 is a deep learning inference engine of model HT108CB100-BTB, the network communication module 6 is a gigabit Ethernet communication module of model ZH_DZ_SW_YT9218NH, and the power management module 7 is a power module of model KUB4812QB-10A.
[0037] The video image splicing and fusion module 4 (model: SOM3559A-A16C0D-A2-C) is manufactured by Beijing ZOHETEC Technology Co., Ltd. The SOM3559A-A16C0D-A2-C is a series model of the company's edge fusion controller, which supports multi-channel video splicing and 4K resolution processing, and is widely used in the field of security monitoring.
[0038] The Video Intelligent Analysis Module 5 (equipped with the HT108CB100-BTB deep learning inference engine) supports human detection, vehicle detection, and drone detection. The core technology provider for the HT108CB100-BTB deep learning inference engine is Beijing Zhuohe Technology Co., Ltd. and the open-source algorithm community (YOLO6 algorithm). The Video Intelligent Analysis Module 5 uses a professional high-performance processor launched by Hisilicon Technology Co., Ltd. for the security monitoring market. The SOC integrates a high-efficiency neural network inference unit with up to 10 TOPS INT8 and supports mainstream neural network frameworks in the industry. It also has a built-in dual-core Vision DSP. Many security manufacturers have integrated this chip solution into their intelligent analysis modules.
[0039] The manufacturer of Network Communication Module 6 (Model: ZH_DZ_SW_YT9218NH) is Beijing ZOHETEC Technology Co., Ltd. This model is an industrial-grade gigabit Ethernet module that supports wide-temperature operation, complies with the IEEE 802.3ab standard, and is commonly used for network transmission in industrial monitoring equipment.
[0040] The power management module 7 (model: KUB4812QB-10A) is manufactured by Guangzhou Mornsun Technology Co., Ltd. (MORNSON). KUB4812QB-10A is a standard model with 24V input and 40W output, featuring low power consumption and high conversion efficiency.
[0041] The multi-lens ring lens assembly, dual-spectrum imaging unit, video image stitching and fusion module 4, video intelligent analysis module 5, network communication module 6, and power management module 7 are all installed in the inner cavity of the housing base 3 through sealing rings and anti-loosening screws to ensure IP67 protection level. The multi-lens ring lens assembly is installed on a unified lens mounting bracket 302, so that the lens optical axis deviation is ≤0.1°.
[0042] The power management module 7 (model KUB4812QB-10A) is powered by 24V, providing stable power (power consumption ≤40W) to ensure continuous operation of all modules. The aluminum alloy housing base 3, detachable top cover 301, and lens mounting bracket 302 of the panoramic housing assembly form a protective and support structure: the machined aluminum alloy lens mounting bracket 302 ensures that the four visible light lenses 1 and four infrared thermal imaging lenses 101 are precisely arranged in a circular array (the angle between adjacent lenses is 45°), with an optical axis deviation ≤0.1°, and the visible light lenses 1 and infrared thermal imaging lenses 101 are located on the same plane, providing a structural accuracy basis for subsequent image stitching; at the same time, the lens mounting bracket 302 uses the thermal conductivity of aluminum alloy to conduct the heat generated by the infrared sensor 202 to the housing base 3, reducing thermal noise and improving infrared imaging stability. In addition, each module is fixed to the inner cavity of the housing base 3 by sealing rings and anti-loosening screws, achieving an IP67 protection level and adapting to harsh outdoor environments;
[0043] Four visible light lenses 1 cover a 180° horizontal field of view and a 30° vertical field of view, transmitting visible light signals to the corresponding visible light sensor 2 (model OS08A20, RGB imaging), which converts them into RGB color image electrical signals.
[0044] Four infrared thermal imaging lenses 101 cover a 180° horizontal field of view and a 40° vertical field of view, transmitting infrared band signals to the corresponding infrared sensor 201 (model RTDS121C, thermal imaging), which converts them into infrared thermal imaging electrical signals.
[0045] The dual-spectral imaging unit forms an independent multi-channel acquisition, simultaneously acquiring visible light and infrared dual-band image data, providing raw data for all-weather monitoring (high-definition visible light imaging during the day and infrared thermal imaging at night);
[0046] The video image stitching and fusion module 4 (model SOM3559A-A16C0D-A2-C, edge fusion controller) receives multi-channel dual-spectrum image data. Based on the precise positional relationship ensured by the lens fixing bracket 302, it stitches and fuses four visible light images to form a 180° horizontal × 30° vertical visible light panoramic video. At the same time, it stitches and fuses four infrared images to form a 180° horizontal × 40° vertical infrared panoramic video. During the fusion process, it supports real-time processing of 4K resolution and adaptive adjustment of the fusion band width to ensure a natural transition at the stitching points without obvious stitching marks, achieving large-area seamless panoramic monitoring.
[0047] The video intelligent analysis module 5 (model HT108CB100-BTB, deep learning inference engine) performs intelligent analysis on the stitched visible light and infrared panoramic video, and uses deep learning algorithms to detect targets such as people, vehicles, and drones in the scene in real time, and outputs detection results (such as target location and type). The network communication module 6 (model ZH_DZ_SW_YT9218NH, gigabit Ethernet) transmits the stitched panoramic video and intelligent analysis results to the PC via gigabit Ethernet to realize real-time image viewing; at the same time, it receives control commands from the PC to complete operations such as adjusting device parameters (such as lens focal length and analysis sensitivity).
[0048] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Additionally, in the accompanying drawings of this utility model, the fill patterns are merely for distinguishing layers and do not constitute any other limitation.
[0049] Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the present invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A non-structural multi-modal arrayed ultra-high definition intelligent opto-electronic detector, characterized in that: The panoramic shell assembly includes a machine shell base (3), a detachable top cover (301), and a lens fixing support (302), the multi-view annular lens group is embedded and fixed on the lens fixing support (302) of the machine shell base (3), the visible light lens (1) above the lens fixing support (302) and the infrared thermal imaging lens (101) below are located in the same plane. The number of the visible light lens (1) and the infrared thermal imaging lens (101) is 4 respectively, the adjacent visible light lens (1) has an angle of 45°, covering a horizontal visual angle of 180° and a vertical visual angle of 30°; the adjacent infrared thermal imaging lens (101) has an angle of 45°, covering a horizontal visual angle of 180° and a vertical visual angle of 40°. The lens fixing support (302) is a machined aluminum alloy support, which is used to ensure the horizontal angle precision of adjacent lenses and the stability of the optical axis, and can conduct heat to the machine shell base (3). The machine shell base (3) and the detachable top cover (301) of the panoramic shell assembly are respectively an aluminum alloy machine shell base and an aluminum alloy detachable top cover.
2. The non-structural multi-modal arrayed ultra-high definition smart optoelectronic detector of claim 1, wherein: The visible light sensor (2) is an RGB imaging sensor of model OS08A20, and the infrared sensor (201) is a thermal imaging sensor of model RTDS121C.
3. The non-structural multi-modal arrayed ultra-high definition smart opto-electronic detector of claim 1, wherein: The video image splicing and fusion module (4) is an edge fusion controller of model SOM3559A-A16C0D-A2-C, the video intelligent analysis module (5) is a deep learning inference engine of model HT108CB100-BTB, the network communication module (6) is a gigabit Ethernet communication module of model ZH_DZ_SW_YT9218NH, and the power management module (7) is a power module of model KUB4812QB-10A.
4. The non-structural multi-modal arrayed ultra-high definition smart optoelectronic detector of claim 1, wherein: The multi-view annular lens group, the dual-spectrum imaging unit, the video image splicing and fusion module (4), the video intelligent analysis module (5), the network communication module (6), and the power management module (7) are all installed in the inner cavity of the machine shell base (3) through a sealing ring and a anti-loosening screw.
5. The non-structural multi-modal arrayed ultra-high definition smart optoelectronic detector of claim 1, wherein: 6. The non-structural multi-modal arrayed ultra-high definition smart optoelectronic detector of claim 1, wherein: 7. The non-structural multi-modal arrayed ultra-high definition smart optoelectronic detector of claim 1, wherein: