Search

CN-122027909-A - Imaging method for partitioning exposure scheduling and glare suppression of vehicle-mounted lens

CN122027909ACN 122027909 ACN122027909 ACN 122027909ACN-122027909-A

Abstract

The invention relates to the technical field of vehicle-mounted optical imaging, in particular to a vehicle-mounted lens partition exposure scheduling and glare suppression imaging method which comprises the steps of S1, acquiring original environment light field distribution information in front of a vehicle, identifying a strong light source area and a conventional area except the strong light source area in a visual field range based on the original light field distribution information, S2, generating a dynamic partition control instruction based on the identified strong light source area, executing visual field segmentation on a photosurface of an imaging sensor according to the dynamic partition control instruction, distributing a first exposure period for the strong light source area and distributing a second exposure period for the conventional area, wherein the duration of the first exposure period is smaller than that of the second exposure period, and fundamentally preventing saturation of a photosurface pixel unit caused by receiving excessive light energy by the fact that the photosurface pixel unit is projected to the center of the photosurface through a deformable lens array.

Inventors

  • TANG HAO
  • LU ZHANJUN
  • LU YANG

Assignees

  • 深圳市永泰光电有限公司

Dates

Publication Date
20260512
Application Date
20260407

Claims (9)

  1. 1. The imaging method for the zoned exposure scheduling and glare suppression of the vehicle-mounted lens is characterized by comprising the following steps of: s1, acquiring original environment light field distribution information in front of a vehicle, and identifying a strong light source area and a conventional area except the strong light source area in a view field range based on the original light field distribution information; s2, generating a dynamic partition control instruction based on the identified strong light source region, and executing field segmentation on a photosurface of an imaging sensor according to the dynamic partition control instruction, distributing a first exposure period for the strong light source region and a second exposure period for the conventional region, wherein the duration of the first exposure period is smaller than that of the second exposure period; Acquiring state parameters of a deformable lens array, attaching the deformable lens array to a photosurface surface layer of an imaging sensor, and determining a first lens unit subset needing to change curvature and a second lens unit subset keeping a planar light transmission state in the deformable lens array based on the shape and the position of a strong light source area; Applying a first driving voltage to the first lens unit subset to deform the first lens unit subset to change a focusing light path of incident light, so that strong light spots originally projected on a photosensitive surface are offset or scattered, and the second lens unit subset is kept in an original light transmission state to keep normal imaging of a conventional area; the first exposure period is allocated to the strong light source area and is executed based on the deformable lens array, a light deflection path corresponding to the first lens unit subset after deformation is determined, and whether the deflected light is projected to a special extinction charge storage area preset on a photosurface or not is judged based on the light deflection path; S3, after the first exposure period and the second exposure period are finished, respectively reading first image information corresponding to the strong light source area and second image information corresponding to the conventional area; And S4, synthesizing the first image information and the second image information based on the image fusion weight to generate a full-view field image.
  2. 2. The method for zoned exposure scheduling and glare suppression imaging of a vehicle lens according to claim 1, wherein in S1, the original ambient light field distribution information in front of the vehicle is obtained by an auxiliary optical sensor disposed outside the main light path of the lens, the auxiliary optical sensor comprises an independent pixel array, and full-field light intensity data of a front scene is continuously collected at a sampling frequency lower than that of the main imaging sensor; Each pixel unit in the pixel array converts the received optical signals into electric signals, light intensity values corresponding to pixel positions are generated after analog-to-digital conversion, the light intensity values of all pixels form a preliminary light intensity distribution matrix, and row-column coordinates of the preliminary light intensity distribution matrix are in one-to-one correspondence with the space positions of the field of view and are used for representing original energy distribution of the environmental light field on the light sensing surface.
  3. 3. The method for on-vehicle lens partition exposure scheduling and glare suppression imaging according to claim 2, wherein when identifying a strong light source region based on original ambient light field distribution information, first traversing each pixel unit in a preliminary light intensity distribution matrix, and comparing a light intensity value of each pixel unit with a preset first threshold; The preset first threshold value is set according to the saturation response value of the main imaging sensor, and when the pixel light intensity value exceeds the preset first threshold value, the pixel point is indicated to possibly be subjected to strong light irradiation, so that the pixel is marked as a candidate highlight point; after full graph traversal is completed, all spatially adjacent candidate highlight points are subjected to connected domain analysis, adjacent candidate highlight points are aggregated into one or more continuous pixel unit sets through neighborhood connectivity judgment, each set is a candidate region, and the candidate region represents the range of a potential strong light source.
  4. 4. The method for vehicle-mounted lens partition exposure scheduling and glare suppression imaging according to claim 3, wherein for each candidate region, calculating a geometric center coordinate and an edge diffusion gradient of the candidate region, wherein the geometric center coordinate is obtained by respectively averaging the abscissa and the ordinate of all pixels in the candidate region and is used for locating the center position of a strong light source, and the edge diffusion gradient is obtained by analyzing the change rate of the light intensity value of the pixels at the boundary of the candidate region; Fitting a continuous and closed boundary contour based on the geometric center coordinates and the edge diffusion gradient, and radially searching critical points with the light intensity values falling below a first threshold value along a plurality of directions by taking the geometric center as a starting point to serve as an initial boundary point set; then screening and adjusting the initial boundary point according to the edge diffusion gradient, wherein the point with a large gradient value shows that the boundary is steep, the point is directly reserved, the point with a small gradient value shows that gradual transition exists, and a certain distance is required to be expanded outwards to contain the transition zone; and finally, adopting an interpolation algorithm or a curve fitting algorithm to smoothly connect the adjusted boundary point sets to generate a continuous and closed curve, determining an internal area surrounded by the curve as a strong light source area, and identifying all areas outside the closed curve in the field of view as conventional areas.
  5. 5. The method for zoned exposure scheduling and glare suppression imaging of a vehicle-mounted lens according to claim 1, wherein the method for zoned exposure scheduling and glare suppression imaging of a vehicle-mounted lens is characterized by judging whether the deflected light is projected onto a dedicated extinction charge storage area preset on a photosurface based on a light deflection path, and specifically comprises the following steps: Acquiring curvature parameters and inclination angle parameters of each deformed lens unit in the first lens unit subset, calculating an emergent direction vector of each ray passing through the deformable lens unit by combining an original direction vector of the incident ray through a ray tracing algorithm, and calculating theoretical falling point coordinates of each ray on a photosurface according to the emergent direction vector and the vertical distance between the lens unit subset and the photosurface to generate a falling point distribution set containing falling point coordinates of all deflected rays; Reading a boundary coordinate range of a special extinction charge storage area preset in an imaging sensor, wherein the boundary coordinate range comprises a transverse initial coordinate, a transverse end coordinate, a longitudinal initial coordinate and a longitudinal end coordinate of the storage area in a photosurface two-dimensional coordinate system, and forming a plurality of matrixes or polygonal closed areas by the coordinate values to limit the physical position of the special extinction charge storage area on the photosurface; Comparing each drop point coordinate in the drop point distribution set with the boundary coordinate range of the special extinction charge storage area one by one, judging whether the transverse numerical value of the drop point coordinate is positioned between the transverse initial coordinate and the transverse termination coordinate, and judging whether the longitudinal numerical value of the drop point coordinate is positioned between the longitudinal initial coordinate and the longitudinal termination coordinate; When the transverse numerical value and the longitudinal numerical value of a certain drop point coordinate are simultaneously satisfied and located in the corresponding boundary coordinate range, the light corresponding to the drop point coordinate is judged to be projected into the special extinction charge storage area, the drop point coordinate is marked as an effective projection point, and when any one of the transverse numerical value or the longitudinal numerical value of the certain drop point coordinate is not satisfied and located in the corresponding boundary coordinate range, the light corresponding to the drop point coordinate is judged not to be projected into the special extinction charge storage area; And generating a corresponding control signal according to the judgment result, generating an enabling signal to be sent to a charge reading circuit of the special extinction charge storage area when the special extinction charge storage area is judged to be projected, starting a charge reading operation in a first exposure period to acquire the light energy of a strong light source area, and generating an adjusting signal to be fed back to a control unit of the deformable lens array when the special extinction charge storage area is judged not to be projected so as to correct assignment or waveform of the first driving voltage.
  6. 6. The in-vehicle lens partition exposure scheduling and flare suppression imaging method according to claim 1, wherein in S3, the reading operation of the first image information is performed based on the arrival of the end time of the first exposure period at the integration completion signal of the dedicated extinction charge storage area and the reading operation of the second image information is performed based on the arrival of the end time of the second exposure period at the regular imaging area, wherein the end time of the first exposure period is defined by the integration completion signal of the dedicated extinction charge storage area and the end time of the second exposure period is defined by the integration completion signal of the regular imaging area; The reading operation of the first image information converts the charge signal into the voltage signal by activating the row gating circuit and the column gating circuit corresponding to the special extinction charge storage area, and the voltage signal is generated by combining the digital gray values according to the pixel arrangement sequence through the analog-to-digital converter, and the reading operation of the second image information converts the charge signal into the voltage signal by activating the row gating circuit and the column gating circuit corresponding to the conventional imaging area, and the voltage signal is generated by combining the digital gray values according to the pixel arrangement sequence through the analog-to-digital converter.
  7. 7. The method for on-vehicle lens partition exposure scheduling and glare suppression imaging according to claim 6, wherein in the process of reading the first image information and the second image information, the timing synchronization generator continuously outputs a synchronization clock signal to the reading circuit of the dedicated extinction charge storage area and the reading circuit of the regular imaging area to ensure that the two reading operations are kept accurately synchronized on a time axis, and after the reading of the first image information and the second image information is completed, the two image information are respectively stored in two independent buffer areas of the image data fusion unit, and a data ready state signal is output to a control node of the image data fusion unit to start a subsequent image synthesis flow.
  8. 8. The method for vehicle-mounted lens partition exposure scheduling and glare suppression imaging according to claim 1, wherein in S4, a preset image fusion weight corresponding to the spatial distribution of the strong light source region and the regular region is obtained, wherein the fusion weight coefficient corresponding to the strong light source region is smaller than the fusion weight coefficient corresponding to the regular region; Performing spatial position registration operation on the first image information and the second image information, and establishing a one-to-one correspondence between each pixel point in the first image information and a corresponding pixel point in the second image information according to a pre-calibrated coordinate mapping relation, wherein the coordinate mapping relation is generated based on relative position coordinates of a special extinction charge storage area and a conventional imaging area on a photosurface and deflection parameters of a deformable liquid lens array on an optical path; For each registered pixel position, acquiring a first gray value from the first image information, acquiring a second gray value from the second image information, acquiring a corresponding first weight coefficient and second weight coefficient from the fusion weight matrix, multiplying the first gray value by the first weight coefficient to obtain a first weighted value, multiplying the second gray value by the second weight coefficient to obtain a second weighted value, and adding the first weighted value and the second weighted value to obtain the synthesized gray value of the pixel position. The method for vehicle-mounted lens partition exposure scheduling and glare suppression imaging according to claim 1, wherein in S4, a preset image fusion weight corresponding to the spatial distribution of the strong light source region and the regular region is obtained, wherein the fusion weight coefficient corresponding to the strong light source region is smaller than the fusion weight coefficient corresponding to the regular region; Performing spatial position registration operation on the first image information and the second image information, and establishing a one-to-one correspondence between each pixel point in the first image information and a corresponding pixel point in the second image information according to a pre-calibrated coordinate mapping relation, wherein the coordinate mapping relation is generated based on relative position coordinates of a special extinction charge storage area and a conventional imaging area on a photosurface and deflection parameters of a deformable liquid lens array on an optical path; For each registered pixel position, acquiring a first gray value from the first image information, acquiring a second gray value from the second image information, acquiring a corresponding first weight coefficient and second weight coefficient from the fusion weight matrix, multiplying the first gray value by the first weight coefficient to obtain a first weighted value, multiplying the second gray value by the second weight coefficient to obtain a second weighted value, and adding the first weighted value and the second weighted value to obtain the synthesized gray value of the pixel position.
  9. 9. The method for zoned exposure scheduling and glare suppression imaging of a vehicle-mounted lens according to claim 8, wherein boundary smoothing is performed on pixels at the junction of a strong light source region and a conventional region, boundary contours of the strong light source region are identified, and pixels located on the boundary contours and within a preset neighborhood range inside and outside the boundary contours are replaced with gradient fusion weights instead of fixed fusion weights, the gradient fusion weights are linearly interpolated from the boundary contours to the inner side and the outer side to generate a smooth transition of weight coefficients from the strong light source region side to the conventional region side, and composite gray values of the pixels are recalculated based on the gradient fusion weights; And executing dynamic range compression and output operation on the synthesized gray values of all pixels, determining a mapping curve according to the input dynamic range of the target display equipment or the subsequent image processing module, integrally mapping the gray values of all pixels in the synthesized image according to the mapping curve, keeping the relative relation between highlight details of a strong light source area and dark part details of a conventional area unchanged in the mapping process, and finally generating a full-view field image with high dynamic range.

Description

Imaging method for partitioning exposure scheduling and glare suppression of vehicle-mounted lens Technical Field The invention relates to the technical field of vehicle-mounted optical imaging, in particular to a vehicle-mounted lens partition exposure scheduling and glare suppression imaging method. Background The vehicle-mounted vision system is a core component for realizing environment perception by the auxiliary driving and automatic driving functions, and in the running process of a vehicle, the imaging lens is required to cope with complex and changeable illumination environments, such as severe illumination changes caused by tunnel entrances and exits, direct irradiation of a high beam of the opposite vehicle, low-angle strong light generated by sunset time division, specular reflection generated by a water depression on a road surface and the like, and the extreme illumination scenes can cause light energy received by a local area of a vehicle-mounted camera sensor to far exceed the saturated capacity of the local area, so that overexposure spots and internal reflection glare are formed. In order to cope with strong light interference, the prior art mainly shortens the whole exposure time through an electronic shutter, but can inhibit overexposure, but simultaneously sacrifices the signal to noise ratio of a dark area scene, so that the details of the dark area are lost, the method for adjusting the overall light passing amount through a mechanical aperture or a liquid lens is still the whole dimming, the imaging requirements of a high light area and a dark area cannot be met in a single frame, and in addition, the traditional digital image restoration technology, such as the pixel interpolation of the high light area, can only guess lost information through an algorithm, but cannot restore the real physical details. Disclosure of Invention Aiming at the technical problems in the prior art, the invention provides a vehicle-mounted lens partition exposure scheduling and glare suppression imaging method. The technical scheme for solving the technical problems is as follows, the vehicle-mounted lens partition exposure scheduling and glare suppression imaging method comprises the following steps: s1, acquiring original environment light field distribution information in front of a vehicle, and identifying a strong light source area and a conventional area except the strong light source area in a view field range based on the original light field distribution information; s2, generating a dynamic partition control instruction based on the identified strong light source region, and executing field segmentation on a photosurface of an imaging sensor according to the dynamic partition control instruction, distributing a first exposure period for the strong light source region and a second exposure period for the conventional region, wherein the duration of the first exposure period is smaller than that of the second exposure period; s3, after the first exposure period and the second exposure period are finished, respectively reading first image information corresponding to the strong light area and second image information corresponding to the normal area; And S4, synthesizing the first image information and the second image information based on the image fusion weight to generate a full-view field image. In a preferred embodiment, in the step S1, the original ambient light field distribution information in front of the vehicle is obtained by an auxiliary optical sensor disposed outside the main optical path of the lens, the auxiliary optical sensor comprises an independent pixel array, and the full-field light intensity data of the front scene is continuously collected at a sampling frequency lower than that of the main imaging sensor; each pixel unit in the pixel array converts the received optical signals into electric signals, light intensity values corresponding to pixel positions are generated after analog-to-digital conversion, the light intensity values of all pixels form a preliminary light intensity distribution matrix, and row-column coordinates of the preliminary light intensity distribution matrix are in one-to-one correspondence with space positions of a field of view and are used for representing original energy distribution of an environmental light field on a light sensing surface; When a strong light source area is identified based on original environment light field distribution information, firstly traversing each pixel unit in a preliminary light intensity distribution matrix, and comparing the light intensity value of each pixel unit with a preset first threshold value; The preset first threshold value is set according to the saturation response value of the main imaging sensor, and when the pixel light intensity value exceeds the preset first threshold value, the pixel point is indicated to possibly be subjected to strong light irradiation, so that the pixel is marked as a candidate