Night vision image processing method
By combining a CMOS rolling shutter and an alternating fill light, the exposure mode and brightness of the night view image are adjusted, solving the problem of local overexposure and underexposure in night doorbell devices. This achieves uniform image brightness and reduced power consumption, improving the battery life and image quality of the battery doorbell.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-05
- Publication Date
- 2026-04-07
AI Technical Summary
In existing technologies, high exposure of doorbell devices at night leads to overexposure of wall areas, while reducing exposure results in excessive darkness of corridor areas. Furthermore, dual-frame/multi-frame fusion technology increases device power consumption, affecting battery life and image quality.
The night vision image is adjusted by using a CMOS rolling shutter line-by-line exposure method, combined with an alternating fill light. The fill light is adjusted by adjusting the fill light amount through a column-by-column exposure mode, and the exposure brightness is processed in different areas to achieve local brightness uniformity and reduce the overall power consumption of the device.
It effectively improves the dynamic range of night vision images, reduces device power consumption, enhances the battery life of the doorbell, and improves image brightness uniformity and the processing effect of moving objects.
Smart Images
Figure CN121815083A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of image processing technology, and in particular to a method for processing night vision images. Background Technology
[0002] The use of outdoor doorbells is becoming increasingly widespread both domestically and internationally, and the field of view of doorbells is also showing a trend of gradual increase. However, due to the diversity of home environments, many users have walls at their doorsteps. When the doorbell's supplementary light is turned on at night, a certain proportion of the wall in the image will be illuminated. Since the wall is very close to the doorbell and has good reflective properties, the wall area appears severely overexposed under the high exposure of the doorbell device at night. To suppress this overexposure, the image exposure is usually reduced, which results in the hallway area, which needs to be focused on, being too dark.
[0003] To address this type of image display where some areas are too bright and others are too dark, digital image technology typically employs optical wide dynamic range processing, using long and short exposure fusion techniques to achieve the effect of not overexposing bright areas while simultaneously increasing the brightness of dark areas.
[0004] Optical wide dynamic range (WDR) technology, as a relatively mature technology for addressing insufficient dynamic range, is widely used. However, achieving WDR first requires high-performance chips; for example, to achieve 20 frames per second (fps), the sensor needs to provide 40 fps data streams, and the main controller needs to support long and short frame fusion. Secondly, continuous dual-frame / multi-frame fusion processing leads to a significant increase in device power consumption, resulting in increased energy consumption and heat generation. In battery-powered cameras, this significantly reduces battery life; in cameras with constant power supply, it introduces thermal noise that further affects image quality. Finally, dual-frame / multi-frame fusion technology performs well in clearly defined overbright and underbright areas, but it produces unnatural brightness in areas with intermediate brightness. Furthermore, processing moving objects in these areas results in severe motion blur and poor brightness gradation. Therefore, dual-frame / multi-frame fused WDR has limited applicability in scenarios with objects moving at close range. Summary of the Invention
[0005] This invention addresses the significant increase in power consumption in existing devices, leading to increased energy consumption and heat generation. In battery-powered cameras, this results in a substantial decrease in battery life, while in cameras powered by constant power, it causes additional thermal noise that negatively impacts image quality. Furthermore, while dual-frame / multi-frame fusion technology performs well in clearly defined overly bright and dark areas, it produces unnatural brightness in areas with intermediate brightness. Additionally, it causes severe motion blur and poor brightness gradation when processing moving objects in these areas. This invention provides a method for processing night vision images.
[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: A method for processing night vision images, used in a doorbell monitoring device, the method comprising: Adjusting the exposure mode for night view images: The exposure mode for night view images is adjusted using the CMOS rolling shutter. Exposure brightness processing: This involves processing the exposure brightness of night view images after adjusting the exposure mode.
[0007] As a preferred option: the exposure mode of the night view image is adjusted by rotating the CMOS by 90 degrees using the CMOS rolling shutter line-by-line exposure method to change the line-by-line exposure mode of the night view image to the column-by-column exposure mode of the night view image.
[0008] As a preferred option, the exposure brightness processing includes: Image mode switching is achieved by switching camera modes to obtain the night view image after the switch. The night view image area is divided into n regions for the night view image after the camera mode is switched. Exposure brightness data is obtained by calculating the exposure brightness data for the night view image area divided into n regions using the supplementary light brightness modulation curve. The night view image is adjusted based on the exposure brightness data to obtain a night view image with adjusted exposure brightness.
[0009] As a preferred option: switch the image mode, switch the camera from day mode to night mode, turn on the infrared fill light, convert the color image of the camera to a black and white image, and switch the infrared filter to a white glass filter.
[0010] As a preferred method, the acquisition of exposure brightness data includes: acquiring the exposure response factor; and acquiring exposure brightness data by combining the supplementary light brightness modulation curve with the exposure response factor.
[0011] As a preferred method, the acquisition of exposure brightness data includes: acquiring the exposure response factor K(i); and acquiring exposure brightness data by combining the supplementary light brightness modulation curve with the exposure response factor K(i).
[0012] As a preferred method, the exposure response factor K(i) is obtained by acquiring a night vision image of any frame after the camera mode has been switched to night vision mode and supplemented with peak DC mode illumination, and then calculating the exposure response factor K(i). ; Where i is a positive integer from 1 to n; P1(i) is the average brightness of the night view image in this frame; E(1) is the exposure of the night view image in this frame; and I0 is the set peak DC supplementary light energy. .
[0013] As a preferred embodiment, the acquisition of exposure brightness data I(i) includes: setting the desired brightness value of the night view image to A; Exposure brightness data I(i) is calculated for the night view image regions divided into n areas using the supplementary light brightness modulation curve. ; Where i is a positive integer from 1 to n; A is the expected brightness value of the night view image; K(i) is the exposure response factor of the i-th region; E is the exposure of the current frame, and E(1) is the exposure of the night view image in that frame.
[0014] To address the aforementioned technical problems, the present invention also provides a night vision image processing system, comprising: The night view image exposure mode adjustment module adjusts the exposure mode of the night view image through the CMOS rolling shutter; The exposure brightness processing module processes the exposure brightness of night view images after the exposure mode has been adjusted.
[0015] As a preferred option: the night view image exposure mode adjustment module adjusts the night view image from the line-by-line exposure mode to the column-by-column exposure mode by rotating the CMOS by 90 degrees through the CMOS rolling shutter line-by-line exposure method.
[0016] As a preferred option, the exposure brightness processing includes: Image mode switching is achieved by switching camera modes to obtain the night view image after the switch. The night view image area is divided into n regions for the night view image after the camera mode is switched. Exposure brightness data is obtained by calculating the exposure brightness data for the night view image area divided into n regions using the supplementary light brightness modulation curve. The night view image is adjusted based on the exposure brightness data to obtain a night view image with adjusted exposure brightness.
[0017] The present invention has significant technical effects due to the adoption of the above technical solutions.
[0018] This invention utilizes a CMOS rolling shutter exposure method, combined with an alternating fill light, to physically attenuate the fill light in overexposed areas on one side, thereby eliminating the difference in fill light intensity between the wall and walkway areas near the equipment.
[0019] This invention utilizes a CMOS rolling shutter line-by-line exposure method, rotating the CMOS by 90 degrees to achieve column-by-column exposure of the image. It analyzes the convergence process of image brightness after the fill light is turned on in night mode to determine the relationship between the brightness of different columns in the image and the exposure. By controlling the fill light brightness of each frame when the corresponding column is exposed, the difference in fill light intensity in different areas of the overall image is greatly reduced, avoiding abnormal image performance of local overbrightness and local underbrightness. This achieves improved image brightness dynamic range while reducing overall power consumption.
[0020] The method designed in this invention has no additional requirements on the performance of CMOS and main control. At the same time, the reasonable distribution of supplementary light energy can achieve effective supplementary light and improve dynamic range, while saving the extra supplementary light energy that causes polarization of image brightness and reducing the power consumption of the whole machine.
[0021] The method designed in this invention for battery-powered doorbells not only solves the problem of localized dark images caused by wall reflections, but also significantly reduces the overall power consumption of the device and increases its battery life. Attached Figure Description
[0022] Figure 1 This diagram illustrates the difference between rolling shutter and global shutter exposure modes.
[0023] Figure 2 This is a schematic diagram illustrating how the brightness of the fill light is modulated to achieve weak fill light for the left exposure column.
[0024] Figure 3 This is a schematic diagram of image region division according to the present invention.
[0025] Figure 4 This is a schematic diagram of the exposure response factor curve and the corresponding fill light brightness curve.
[0026] Figure 5-1 These are standard supplementary images.
[0027] Figure 5-2 This is an image showing the effect after reducing the lighting on the right wall.
[0028] Figure 6 A flowchart of the present invention. Detailed Implementation
[0029] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.
[0030] Example 1
[0031] A method for processing night vision images, used in doorbell monitoring devices. Figure 6 The methods include: Adjusting the exposure mode for night view images: The exposure mode for night view images is adjusted using the CMOS rolling shutter. Exposure brightness processing: This involves processing the exposure brightness of night view images after adjusting the exposure mode.
[0032] The night view image exposure mode is adjusted by rotating the CMOS by 90 degrees using the CMOS rolling shutter line-by-line exposure method to change the night view image from line-by-line exposure mode to column-by-column exposure mode.
[0033] The processing of exposure brightness includes: Image mode switching is achieved by switching camera modes to obtain the night view image after the switch. The night view image area is divided into n regions for the night view image after the camera mode is switched. Exposure brightness data is obtained by calculating the exposure brightness data for the night view image area divided into n regions using the supplementary light brightness modulation curve. The night view image is adjusted based on the exposure brightness data to obtain a night view image with adjusted exposure brightness.
[0034] The image mode switching switches the camera from day mode to night mode, turns on the infrared fill light, converts the color image of the camera to a black and white image, and switches the infrared filter to a white glass filter.
[0035] The acquisition of exposure brightness data includes: acquiring the exposure response factor; and acquiring exposure brightness data by combining the supplementary light brightness modulation curve with the exposure response factor.
[0036] The exposure response factor K(i) is obtained by acquiring any frame of night vision image after the camera mode is switched to night vision mode and supplemented with peak DC mode illumination, and then calculating the exposure response factor K(i). ; Where i is a positive integer from 1 to n; P1(i) is the average brightness of the night view image in this frame; E(1) is the exposure of the night view image in this frame; and I0 is the set peak DC supplementary light energy. .
[0037] The acquisition of exposure brightness data I(i) includes: setting the desired brightness value of the night view image to A; Exposure brightness data I(i) is calculated for the night view image regions divided into n areas using the supplementary light brightness modulation curve. ; Where i is a positive integer from 1 to n; A is the expected brightness value of the night view image; K(i) is the exposure response factor of the i-th region; E is the exposure of the current frame, and E(1) is the exposure of the night view image in that frame.
[0038] Example 2
[0039] Based on Embodiment 1, this embodiment is a night vision image processing system, which is implemented by the method described in the claims, comprising: The night view image exposure mode adjustment module adjusts the exposure mode of the night view image through the CMOS rolling shutter; The exposure brightness processing module processes the exposure brightness of night view images after the exposure mode has been adjusted.
[0040] The night view image exposure mode adjustment module adjusts the night view image from a line-by-line exposure mode to a column-by-column exposure mode by rotating the CMOS by 90 degrees using a CMOS rolling shutter line-by-line exposure method.
[0041] The processing of exposure brightness includes: Image mode switching is achieved by switching camera modes to obtain the night view image after the switch. The night view image area is divided into n regions for the night view image after the camera mode is switched. Exposure brightness data is obtained by calculating the exposure brightness data for the night view image area divided into n regions using the supplementary light brightness modulation curve. The night view image is adjusted based on the exposure brightness data to obtain a night view image with adjusted exposure brightness.
[0042] Example 3
[0043] Figure 1 This demonstrates the differences between the rolling shutter exposure method commonly used in CMOS sensors and the global shutter exposure method commonly used in CCD sensors. It also shows how rotating the CMOS sensor by 90° converts line-by-line exposure to column-by-column exposure. This allows control of the brightness of each exposed row by adjusting the brightness of the supplementary light during the exposure of each column of the CMOS sensor. Figure 2 As shown.
[0044] Overexposure of walls in doorbell cameras at night is often more pronounced on walls at close range, while overexposure on distant walls is less noticeable. By analyzing multiple frames of the black-and-white night image exposure adjustment process after the device switches from day mode to night mode (color image to black-and-white, infrared fill light turned on, infrared filter switched to white glass filter), the exposure response characteristics of overexposed and underexposed areas can be obtained. Figure 3 As shown, the exposure response factor is obtained by acquiring any frame of a night vision image after switching the camera mode to night vision mode and using peak DC mode illumination; the exposure response factor K(i) is calculated by acquiring any frame of a night vision image after switching the camera mode to night vision mode and using peak DC mode illumination. ; Where i is a positive integer from 1 to n; P1(i) is the average brightness of the night view image in this frame; E(1) is the exposure of the night view image in this frame; and I0 is the set peak DC supplementary light energy. .
[0045] The set peak DC supplemental light energy (with the same fixed maximum supplemental light intensity for each area) is generally achieved by turning on the lights first and then switching between color and black-and-white modes. Therefore, the energy of the supplemental light is already stable, and it is assumed that the system is already in a stable peak energy uniform supplemental light state when entering black-and-white night vision mode.
[0046] The acquisition of exposure brightness data includes: setting the desired brightness value A for the night view image; and calculating the exposure brightness data I(i) for the night view image regions divided into n areas using the supplementary light brightness modulation curve. ; Where i is a positive integer from 1 to n; A is the expected brightness value of the night view image; K(i) is the exposure response factor of the i-th region; E is the exposure of the current frame, and E(1) is the exposure of the night view image in that frame. Figure 4 The distribution of the exposure response factor K(i) and the corresponding exposure brightness data I(i) curve in the i-th region is described. Figure 5-1 and 5-2 In the two comparison images, Figure 5-2 The middle image shows the effect after the fill light on the right wall is reduced.
Claims
1. A method for processing night vision images, used in a doorbell monitoring device, characterized in that, The methods include: Adjusting the exposure mode for night view images: The exposure mode for night view images is adjusted using the CMOS rolling shutter. Exposure brightness processing: Exposure brightness processing is performed on the night image after the exposure mode is adjusted; The exposure mode of the night image is adjusted by rotating the CMOS by 90 degrees to change the line-by-line exposure mode of the night image to column-by-column exposure mode through the CMOS rolling shutter line-by-line exposure mode. Image mode switching is achieved by switching camera modes to obtain the night view image after the switch. The night view image area is divided into n regions for the night view image after the camera mode is switched. Exposure brightness data is obtained by calculating the exposure brightness data of the night view image area after dividing it into n regions using the supplementary light brightness modulation curve; The night view image is adjusted based on the exposure brightness data to obtain a night view image with adjusted exposure brightness.
2. The method for processing night view images according to claim 1, characterized in that: The image mode switching switches the camera from day mode to night mode, turns on the infrared fill light, converts the color image of the camera to a black and white image, and switches the infrared filter to a white glass filter.
3. The method for processing night view images according to claim 1, characterized in that: The acquisition of exposure brightness data includes: acquiring the exposure response factor K(i); and acquiring exposure brightness data by combining the supplementary light brightness modulation curve with the exposure response factor K(i).
4. The method for processing night view images according to claim 1, characterized in that: The exposure response factor K(i) is obtained by acquiring any frame of night vision image after the camera mode is switched to night vision mode and supplemented with peak DC mode illumination, and then calculating the exposure response factor K(i). ; Where i is a positive integer from 1 to n; P1(i) is the average brightness of the night view image in this frame; E(1) is the exposure of the night view image in this frame; and I0 is the set peak DC supplementary light energy. .
5. The method for processing night view images according to claim 1, characterized in that: The acquisition of exposure brightness data I(i) includes: setting the desired brightness value of the night view image to A; Exposure brightness data I(i) is calculated for the night view image regions divided into n areas using the supplementary light brightness modulation curve. ; Where i is a positive integer from 1 to n; A is the expected brightness value of the night view image; K(i) is the exposure response factor of the i-th region; E is the exposure of the current frame, and E(1) is the exposure of the night view image in that frame.
6. A night vision image processing system for use in a doorbell monitoring device, characterized in that, The system implemented by any one of the methods of claims 1-5 includes: The night view image exposure mode adjustment module adjusts the exposure mode of the night view image through the CMOS rolling shutter; The exposure brightness processing module processes the exposure brightness of night view images after the exposure mode has been adjusted.
7. The night vision image processing system according to claim 6, characterized in that, The night view image exposure mode adjustment module adjusts the night view image from a line-by-line exposure mode to a column-by-column exposure mode by rotating the CMOS by 90 degrees using a CMOS rolling shutter line-by-line exposure method.
8. The night vision image processing system according to claim 6, characterized in that, The processing of exposure brightness includes: Image mode switching is achieved by switching camera modes to obtain the night view image after the switch. The night view image area is divided into n regions for the night view image after the camera mode is switched. Exposure brightness data is obtained by calculating the exposure brightness data of the night view image area after dividing it into n regions using the supplementary light brightness modulation curve; The night view image is adjusted based on the exposure brightness data to obtain a night view image with adjusted exposure brightness.