License plate recognition light supplementing method and system based on dual-mode vision and dynamic beam shaping

Abstract

The application belongs to the technical field of intelligent transportation, and discloses a license plate recognition light supplementing method and system based on dual-mode vision and dynamic light beam shaping. The method first uses a near-infrared imaging system invisible to human eyes to continuously and imperceptibly detect vehicles entering a recognition area, and analyzes images in real time to accurately estimate the position and size of a license plate. Then, the system controls a visible light illumination module to generate a customized light beam highly matched with the license plate area in shape and size according to the estimated license plate area coordinates, and performs instantaneous and accurate light supplementing on the area. While supplementing light, a visible light image sensor is synchronously triggered to take a snapshot to obtain a high-definition color license plate image. The application eliminates glare interference from a physical source, guarantees high-quality color image acquisition capability, and improves driving safety and recognition accuracy.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation technology, specifically relating to a license plate recognition supplementary lighting method and system based on dual-mode vision and dynamic beam shaping. Background Technology

[0002] License plate recognition systems in parking lots, highway toll booths, and urban road checkpoints are core technologies for achieving automated traffic management. To ensure recognition rates in low-light environments such as at night, on rainy days, or in underground parking garages, these systems are typically equipped with supplemental lighting.

[0003] Currently, there are two main types of supplemental lighting technologies:

[0004] 1. Infrared supplementary lighting: This method uses infrared LEDs (such as 850nm or 940nm wavelengths) that are invisible to the human eye for supplementary lighting. Its advantage is that it is completely glare-free and will not interfere with the driver. However, its disadvantages are also quite obvious: infrared imaging loses color information, making it impossible to take color photos. Furthermore, for license plates using specific anti-counterfeiting materials or colors (such as green license plates for new energy vehicles), the contrast between the characters and the background decreases sharply under infrared light, leading to difficulties or failures in recognition.

[0005] 2. Visible Light (White Light) Complementary Lighting: This method uses high-brightness white LEDs for constant or flashing supplementary lighting. Its advantage is the ability to acquire high-definition color images, meeting the needs for recognizing information such as license plate color and vehicle body color. However, its fatal flaw is the generation of intense glare. When a vehicle approaches, the high-intensity white light shines directly into the driver's eyes, potentially causing momentary blindness or visual discomfort, severely impacting driving safety and the driving experience, and causing light pollution.

[0006] To address these issues, existing technologies have attempted several improvements, such as replacing constant-on lights with flashlights to shorten glare duration; using diffusers to soften light; and optimizing images through algorithms like Wide Dynamic Range (WDR) and High Light Suppression (HLC) at the camera end. However, these methods do not fundamentally solve the problem: the instantaneous brightness of flashlights remains glaring; diffusers cause light energy loss and light field dispersion, resulting in a significant amount of stray light entering the eyes; and pure algorithmic optimization cannot reduce the actual light intensity reaching the driver's eyes in the physical world. Summary of the Invention

[0007] The present invention aims to at least partially solve the aforementioned technical problems. Therefore, the objective of the present invention is to provide a license plate recognition supplementary lighting method and system based on dual-mode vision and dynamic beam shaping.

[0008] The technical solution adopted in this invention is as follows:

[0009] A license plate recognition supplementary lighting method based on dual-mode vision and dynamic beam shaping includes the following steps:

[0010] a. When a vehicle enters the recognition area, a near-infrared image sensor is used to capture a near-infrared image containing the vehicle in real time under near-infrared illumination conditions;

[0011] b. Analyze the near-infrared image using a processor to detect vehicle targets and estimate the area coordinates of the license plate;

[0012] c. Based on the area coordinates of the license plate, control a visible light illumination module consisting of a visible light LED array and a dynamic holographic light diffuser to generate a customized beam of light that matches the license plate area in shape, size and direction, and provide instantaneous supplementary lighting to the area;

[0013] d. While the visible light illumination module provides supplemental lighting, the visible light image sensor is simultaneously triggered to expose and capture a high-resolution color license plate image;

[0014] e. Recognize and process the captured color license plate image, and then turn off the visible light illumination module.

[0015] Preferably, in step c, controlling the visible light illumination module specifically includes:

[0016] Based on the regional coordinates of the license plate, activate the portion of LED beads in the visible light LED array that correspond to the coordinates;

[0017] Simultaneously, the dynamic holographic light diffuser is controlled to further reshape the light emitted by the activated LED beads, converging it into a light spot that precisely covers the license plate area.

[0018] Preferably, the dynamic holographic light diffuser is based on an electro-controlled liquid crystal layer to achieve dynamic switching of the beam mode.

[0019] Preferably, in step b, the processor runs a deep learning-based vehicle detection algorithm to locate the vehicle outline from the near-infrared image, and estimates the regional coordinates of the license plate based on the vehicle outline and a preset geometric relationship.

[0020] Preferably, the near-infrared supplementary light uses infrared light with a wavelength of 850nm that is invisible to the human eye, and remains constantly lit during vehicle detection or flashes in sync with the frame rate of the near-infrared image sensor.

[0021] A license plate recognition supplementary lighting system based on dual-mode vision and dynamic beam shaping is provided to implement the above method. The system includes:

[0022] The dual-mode imaging module includes a visible light image sensor for capturing color images and a near-infrared image sensor for capturing near-infrared images.

[0023] A near-infrared illumination module is used to provide supplemental light for the near-infrared image sensor;

[0024] A visible light illumination module, used to provide supplemental light for the visible light image sensor, includes a visible light LED array and a dynamic holographic light diffuser;

[0025] The intelligent beam control system is electrically connected to the dual-mode imaging module and the visible light illumination module;

[0026] The intelligent beam control system is configured as follows:

[0027] The near-infrared image is received from the near-infrared image sensor; the near-infrared image is analyzed to determine the target area of ​​the license plate; based on the target area, the visible light illumination module is controlled to generate a visible light beam that mainly covers the target area; and the visible light image sensor is synchronously triggered to capture the image.

[0028] Preferably, the visible light LED array in the visible light illumination module is arranged in an M×N matrix, and the intelligent beam control system achieves coarse adjustment of the beam position by selectively activating one or more LEDs in the matrix.

[0029] Preferably, the dynamic holographic light diffuser is located on the light output path of the visible light LED array, and is used to finely shape the coarsely adjusted light beam to improve the matching degree between the light spot and the license plate shape and the light utilization rate.

[0030] Preferably, the visible light image sensor and the near-infrared image sensor have calibrated and aligned fields of view to ensure that the license plate area coordinates estimated in the near-infrared image can be accurately mapped to the imaging plane of the visible light image sensor.

[0031] Preferably, the intelligent beam control system incorporates an embedded processor with an integrated neural processing unit for efficiently running vehicle and license plate detection algorithms.

[0032] The beneficial effects of this invention are as follows:

[0033] This invention utilizes near-infrared vision for non-invasive detection and positioning. In a mere microsecond or millisecond-level snapshot, a precisely shaped beam of light, covering only the size of the license plate, provides supplementary illumination. The vast majority of the light energy is concentrated on the license plate, with minimal spillage into the driver's cab or field of vision, thus eliminating glare at its physical source.

[0034] This invention, by employing visible light supplemental illumination, can acquire high-definition color images, is compatible with all types of license plates, including those for new energy vehicles, and retains important information such as vehicle body color. Precise supplemental illumination ensures optimal lighting for the license plate area, resulting in high contrast, which is beneficial for subsequent recognition algorithms.

[0035] This invention can automatically adapt to different vehicle models (cars, SUVs, etc. with different license plate heights) and different positions when vehicles enter, dynamically adjusting the light beam. It has a high degree of intelligence and requires no manual intervention. Compared with constant or wide-area flashing auxiliary lights, this invention provides on-demand, zoned, and instantaneous supplementary lighting with extremely low energy consumption, reducing light pollution to the environment. Attached Figure Description

[0036] Figure 1 This is a structural block diagram of the license plate recognition supplementary lighting system based on dual-mode vision and dynamic beam shaping according to the present invention.

[0037] Figure 2 This is a flowchart of the dynamic beam adjustment process of the present invention.

[0038] In the diagram: 100 - Dual-mode imaging module, 101 - Visible light image sensor, 102 - Near-infrared image sensor, 103 - Near-infrared illumination module, 200 - Visible light illumination module, 300 - Intelligent beam control system. Detailed Implementation

[0039] The technical solutions of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0040] It should be understood that, and also noted, in the embodiments, the functions / actions may appear in a different order than those shown in the figures. For example, depending on the functions / actions involved, they may actually be performed substantially concurrently, or sometimes the two figures shown consecutively may be performed in reverse order.

[0041] Reference Figure 1 The present invention provides a license plate recognition supplementary lighting system based on dual-mode vision and dynamic beam shaping, which mainly includes a dual-mode imaging module 100, a near-infrared illumination module 103, a visible light illumination module 200, and an intelligent beam control system 300.

[0042] The dual-mode imaging module 100 integrates two sensors: a visible light image sensor 101 and a near-infrared image sensor 102.

[0043] The visible light image sensor 101 is used to capture high-definition color license plate images. In this embodiment, a 5-megapixel or higher resolution CMOS sensor can be selected, supporting a 30fps frame rate and wide dynamic range (WDR) function to cope with complex lighting conditions such as daytime backlight and shadows.

[0044] Near-infrared image sensor 102 is used to detect vehicles and locate license plates imperceptibly at night or in low light conditions. Its operating wavelength is matched to that of near-infrared illumination module 103, for example, at 850 nm. This sensor does not require high resolution, but it does require high sensitivity.

[0045] These two sensors can be installed side by side, and their fields of view are basically aligned and synchronized through lens design and structural calibration, so that the coordinates in the near-infrared image can be accurately mapped to the coordinate system of the visible light image.

[0046] The near-infrared illumination module 103 serves the near-infrared image sensor 102. It consists of several low-power 850nm infrared LEDs, with a total power typically less than 3W. In night mode, it can remain constantly lit or flash in sync with the frame rate of the near-infrared sensor, providing a continuous, stable, and invisible illumination environment for vehicle detection.

[0047] The visible light illumination module 200 is the core component of this invention, responsible for generating a precise light beam. It consists of a cascaded visible light LED array and a dynamic holographic light diffuser.

[0048] The visible light LED array consists of multiple independent white LED units, arranged, for example, in a 16×16 or 32×32 matrix. Each LED unit has low power (e.g., <2W) and can be independently driven to light up or turn off by an intelligent beam control system 300. By selectively illuminating specific areas in the array (e.g., a 3×8 sub-region), coarse adjustment of the beam position and approximate range can be achieved.

[0049] In this embodiment, the dynamic holographic light diffuser is specifically an electronically controlled liquid crystal spatial light modulator (LC-SLM). The liquid crystal spatial light modulator is located in front of the LED array. By loading a pre-calculated or real-time generated computational hologram (CGH) phase grayscale image, the liquid crystal spatial light modulator performs pixel-level phase modulation on the incident light wavefront, thereby realizing dynamic switching and precise shaping of the beam mode.

[0050] Specifically, a liquid crystal spatial light modulator (LCoS) is a high-precision, programmable diffractive optical element. Its core is a pixel array based on liquid crystal on silicon (LCoS) technology. By loading a computer-generated grayscale image (i.e., a phase map or computational hologram, CGH) into the driving circuit of the LCOS, each pixel receives a different voltage, thereby controlling the deflection angle of the liquid crystal molecules at that pixel. Due to the birefringence effect of the liquid crystal material, liquid crystal molecules at different deflection angles produce different phase delays for the light passing through them. In this way, the LCOS can perform pixel-level, precise phase modulation of the wavefront of incident light.

[0051] Once the intelligent beam control system 300 determines the license plate's position and the desired light spot shape (e.g., a rectangle with an aspect ratio of approximately 4:1), it calls upon or generates a corresponding computational hologram (CGH) in real time and loads it onto the liquid crystal spatial light modulator. The liquid crystal spatial light modulator then uses this hologram to precisely reshape the light from the rear LED array into the target light spot through diffraction and interference principles. Compared to traditional diffusers or static holograms, the liquid crystal spatial light modulator offers extremely high flexibility and speed (response time <10ms), dynamically generating beams of arbitrary shapes, increasing light utilization by over 90%, and reducing stray light overflowing outside the target area to less than 5%.

[0052] The Intelligent Beam Control System 300 is a high-performance embedded system equipped with a SoC chip from brands such as Hisilicon or MSTAR that integrates a Neural Processing Unit (NPU).

[0053] The intelligent beam control system 300 is responsible for:

[0054] 1. Receive and process video streams from near-infrared image sensor 102 in real time.

[0055] 2. Run lightweight deep learning object detection algorithms (such as YOLO series, SSD, etc.) efficiently on NPU to quickly detect vehicle bounding boxes from near-infrared images.

[0056] 3. Based on the vehicle's bounding box and prior knowledge (such as the license plate usually being located in the lower middle part of the front / rear of the vehicle), estimate the approximate area where the license plate may exist, i.e., the region of interest.

[0057] 4. The image coordinates of the ROI are converted into control commands for the visible light illumination module 200 through a pre-calibrated mapping relationship.

[0058] 5. Generate control signals to precisely drive the corresponding LED beads in the LED array, and load the most suitable computational hologram for the liquid crystal spatial light modulator to achieve the best beam shaping effect.

[0059] 6. At the same time as issuing the dimming command, a synchronization trigger signal is sent to the visible light image sensor 101 to ensure that the supplementary light and the camera exposure are perfectly synchronized.

[0060] 7. Receive the color image captured by the visible light image sensor 101, run a high-precision license plate recognition algorithm, and output the final license plate number, color and other information.

[0061] It should be noted that the pre-calibrated mapping relationship is established before the system leaves the factory by photographing the calibration board to create a homography transformation matrix between the near-infrared image coordinate system and the visible light image coordinate system, thereby achieving accurate coordinate mapping.

[0062] Reference Figure 2 The workflow of this invention (taking nighttime as an example) is as follows:

[0063] 1. Vehicle entry. When a vehicle enters the camera's monitoring range, the system is activated. The near-infrared illumination module 103 is operational, providing basic infrared illumination for the scene.

[0064] 2. Near-infrared image capture and vehicle detection. The near-infrared image sensor 102 continuously captures scene images and transmits them to the intelligent beam control system 300. The system continuously executes the vehicle detection algorithm. If no vehicle is detected, it returns to continue capturing images.

[0065] 3. Identify the vehicle and infer the license plate location. Once a vehicle is detected, the intelligent beam control system 300 immediately infers the approximate ROI of the license plate based on the vehicle's position and size.

[0066] 4. Dynamic Beam Generation. Based on the ROI coordinates, the system performs a dual dimming operation: "Activating a portion of the white light area," which means activating the LED beads corresponding to the ROI in the LED array; simultaneously, "holographic dimming" is performed, which means setting the dynamic holographic light diffuser to the optimal mode to finely shape the beam. The entire process is completed within milliseconds.

[0067] 5. Synchronous capture and recognition. Simultaneously with the generation of the light beam, the visible light image sensor 101 is triggered to capture a color license plate image with precise illumination. The image is then sent to the recognition engine for processing.

[0068] 6. Turn off the white light. After recognition is complete, the visible light illumination module 200 immediately turns off, returning to a state without visible light, and waits for the next vehicle to arrive.

[0069] Through the aforementioned system and method, this invention successfully achieves high-quality visible light illumination and recognition of license plates without generating disturbing glare, greatly improving the safety, comfort, and intelligence of automated traffic management systems. This technology is not only applicable to parking lot entrances and exits but can also be extended to all scenarios requiring license plate recognition, such as urban road monitoring and highway checkpoints.

[0070] This invention is not limited to the above-described optional embodiments. Anyone can derive other various forms of products under the guidance of this invention. However, regardless of any changes made in their shape or structure, any technical solution that falls within the scope of the claims of this invention shall be protected by this invention.

Claims

1. A license plate recognition supplementary lighting method based on dual-mode vision and dynamic beam shaping, characterized in that, Includes the following steps: a. When a vehicle enters the recognition area, a near-infrared image sensor is used to capture a near-infrared image containing the vehicle in real time under near-infrared illumination conditions; b. Analyze the near-infrared image using a processor to detect vehicle targets and estimate the area coordinates of the license plate; c. Based on the area coordinates of the license plate, control a visible light illumination module consisting of a visible light LED array and a dynamic holographic light diffuser to generate a customized beam of light that matches the license plate area in shape, size and direction, and provide instantaneous supplementary lighting to the area; The visible light LED array is arranged in an M×N matrix; The dynamic holographic light diffuser is a liquid crystal spatial light modulator. The dynamic holographic light diffuser is located on the light output path of the visible light LED array and achieves dynamic switching of the beam mode based on the electronically controlled liquid crystal layer. The dynamic holographic light diffuser is a liquid crystal spatial light modulator that reshapes light from a visible light LED array into a target light spot, achieving dynamic switching and precise shaping of the beam mode; the control of the visible light illumination module specifically includes: Based on the regional coordinates of the license plate, the intelligent beam control system selectively activates a portion of the LED beads in the visible light LED array that correspond to the coordinates, thereby achieving coarse adjustment of the beam position. Simultaneously, the dynamic holographic light diffuser is controlled to perform pixel-level phase modulation on the incident light wavefront by loading and calculating a hologram. Through the principles of diffraction and interference, the light emitted by the activated LED beads is further shaped and converged into a light spot that precisely covers the license plate area. d. While the visible light illumination module provides supplemental lighting, the visible light image sensor is simultaneously triggered to expose and capture a high-resolution color license plate image; e. Recognize and process the captured color license plate image, and then turn off the visible light illumination module.

2. The method according to claim 1, characterized in that, In step b, the processor runs a deep learning-based vehicle detection algorithm to locate the vehicle outline from the near-infrared image, and estimates the regional coordinates of the license plate based on the vehicle outline and the preset geometric relationship.

3. The method according to claim 1, characterized in that, The near-infrared supplementary light uses infrared light and remains constantly lit during vehicle detection or flashes in sync with the frame rate of the near-infrared image sensor.

Images