19 results about "Computational photography" patented technology

The invention relates to the technical field of multi-camera image processing and computational photography, and discloses an image processing method and system based on four-camera cross-focal-length continuous zooming fusion. The image processing method comprises the steps of system initialization, image acquisition, zoom routing, automatic ROI extraction and tracking, field-of-view cutting and geometric alignment, image fusion and binocular depth recognition. Through systematized multi-camera collaborative design, an innovative mechanism is introduced in key links such as zoom routing, geometric alignment, image fusion and depth recognition, smooth zoom, space consistency, detail fidelity, power consumption optimization and high-quality 3D perception are realized, the imaging quality is improved through the effects, the application scene is expanded, and the application prospect is wide. And a comprehensive solution is provided for mobile photography, AR and intelligent visual systems.

The invention discloses a night enhanced imaging and dynamic denoising method for an automobile data recorder, relates to the technical field of vehicle-mounted image processing and computational photography, and is used for solving the problems of insufficient brightness and too strong noise of night videos of the automobile data recorder under a low-illumination condition. Estimating the zero-mean Gaussian noise intensity of a read link and the Poisson noise intensity increasing along with the brightness in the same frame, converting the zero-mean Gaussian noise intensity and the Poisson noise intensity into brightness equivalent parameters, and introducing a unified control function G as the scale reference of all enhancement and noise reduction links; defining a sharing priori H and carrying out unified constraint in dense optical flow, depth time sequence denoising and local tone mapping; and in combination with the motion mask and the background model, performing differential space-time denoising and proportional write-back, and in cooperation with motion blur detection, spatial variation motion kernel and constrained deconvolution and multi-frame detail recharge, obtaining target brightness and outputting an RGB frame after proportional write-back.

The invention relates to the field of image enhancement, provides a low-light image enhancement method and device, aims to realize low-light image enhancement with interpretability, low complexity and high enhancement performance, and realizes differential enhancement of different brightness intervals by constructing a bipartite function enhancement model and utilizing correction parameters in the bipartite function enhancement model to regulate and control concavity and convexity of a function curve. Meanwhile, parameters are optimized in combination with a genetic algorithm, the defects that a traditional method is poor in enhancement effect and parameters depend on manual adjustment and optimization are overcome, the problems that a deep learning method is high in complexity and weak in interpretability are solved, core indexes such as detail reservation, color naturalness, real-time performance and scene adaptability of the low-light image are obviously improved, and the method is suitable for popularization and application. The method can be widely applied to the fields of video monitoring, automatic driving, medical imaging, computational photography, industrial visual inspection and the like.

The application provides a single-pixel diffraction imaging method and device, and belongs to the field of coherent diffraction imaging and computational photography, wherein the method comprises the following steps: (1) in the single-pixel imaging light path, different modulation sequences are used to modulate the illumination light beam of the target scene, and a single-pixel detector is used to record the intensity value of a one-dimensional sequence; (2) the intensity value corresponding to the different modulation sequences is used to replace the amplitude value of the reconstructed diffraction light field on the detector plane, and then alternating iteration calculation is performed until the target scene is reconstructed. The application has the advantages of simple structure, strong operability, wide application range and strong expandability. Through the application, the cost of single-pixel diffraction imaging can be reduced, fast and accurate measurement of the target scene is realized, and convenience is brought to high-quality single-pixel imaging.

The invention provides a dark field image acquisition method and device, an image correction method and device, equipment and a medium, and relates to the field of image processing, in particular to the technical fields of digital image signal processing, sensor noise calibration, computational photography and the like. According to the implementation scheme, the method comprises the steps of S1, obtaining a target image generated by an image sensor based on imaging parameters and a shooting part in a first state; s2, on the basis of at least one of the parameter information of the shooting part or the target image, determining whether a dark field acquisition condition is satisfied; and S3, in response to a dark field acquisition condition, switching the shooting part into a second state, acquiring a dark field image under the current imaging parameter, and adding the dark field image into a dark field database.

The application discloses a kind of computational photography method and device, it is related to the field of computational imaging and image processing, in low-illumination environment, realize low cost, high-quality imaging.The scheme includes: before image sensor, configure chopping module, when first area non-equal interval period of chopping module rotation carries out light transmission or light shielding;During imaging, control the rotation of chopping module, so that the starting time of first area is synchronized with the frame exposure of image sensor;After first area leaves image sensor, obtain the first image output by image sensor;Based on the encoding sequence corresponding to first area, determine the point spread vector of each pixel in the to-be-recovered area of first image;Based on the point spread vector of each pixel in the to-be-recovered area of first image, the to-be-recovered area of first image is decoded to obtain clear second image.The described computational photography method and device can be used to high-quality snapshot in low-illumination scene for high-speed moving object.

This application discloses a low-light image compression method and system based on an intelligent diffusion model, relating to the fields of image compression and computational photography. The method includes: mapping a low-light image to a low-dimensional latent space using an encoder to obtain a latent feature representation; performing adaptive quantization on the latent feature representation to obtain a discretized compressed feature representation; using the compressed feature representation as a condition, performing an iterative denoising process through a conditional diffusion model to reconstruct a clear latent feature representation; and converting the reconstructed clear latent feature representation to pixel space using a decoder to generate a reconstructed image. The parameters of the conditional diffusion model and encoder are optimized based on an intelligent gain criterion, which is the ratio of information gain to system complexity. This application achieves efficient compression and high-quality reconstruction of low-light images at extremely low bit rates, effectively suppressing noise and restoring details.

The present application relates to the field of image enhancement, in order to realize the low light image enhancement with explainability, low complexity and high enhancement performance, a low light image enhancement method and device are provided, by constructing a bifunctional enhancement model, using the concave and convex of the correction parameter control function curve to realize the differentiated enhancement of different brightness intervals, and combining with the genetic algorithm to optimize the parameters, both the defects of traditional method such as poor enhancement effect and parameter dependence on artificial optimization are overcome, and the problems of deep learning method such as high complexity and weak explainability are solved, the core indicators such as detail retention, color naturalness, real-time performance and scene adaptability of low light image are obviously improved, and the present application can be widely applied in video monitoring, automatic driving, medical imaging, computational photography, industrial vision detection and other fields.

This invention discloses a distributed defocus stereo camera, relating to the fields of computational photography and depth measurement technology. Existing photogrammetry systems primarily obtain scene depth by acquiring sharp images or image stacks. Depth obtained using sharp images is often relative and lacks accuracy. Methods using focus stacks are complex, slow, costly, and impractical. This invention provides a low-cost, simple operation for acquiring dense and accurate depth, effectively solving this problem. The camera includes a distributed defocus acquisition device, a computational control system, and a depth measurement method. The acquisition device includes a beam splitter, a lens group, an aperture, a focusing device, a ranging device, and a photosensitive imaging device, used to acquire a sharp image, two blurred images, and the focusing distance of the scene. The computational control system calculates the scene depth using the scene data acquired by the distributed defocus acquisition device according to the corresponding depth measurement method.

An endoscopy system having a low-profile multi-imager endoscope. The system is capable of using computational photography to provide enhanced output images using techniques such as super-resolution, foveation, magnification, and two-dimensional to three-dimensional conversion. The enhanced output images can improve clinical decision making and patient treatment. Signals from multiple imagers may be used to affect / adjust handing characteristics of the endoscope or direct semi-robotic guidance thereof.

A method including receiving an image as a portion of a real-world space, placing an anchor on the image, determining a position of the anchor, determine a depth associated with the position of the anchor, applying an image editing algorithm based on the depth to the captured image, and rendering the edited image.

The method of snapshot compressive imaging three-dimensional Gaussian splatting reconstruction belongs to the field of computer vision and computational photography. A reconstruction network based on the hybrid architecture of stochastic gradient Langevin dynamics (SGLD) and explicit three-dimensional Gaussian representation is built. Then, multiple physical perception modules are combined to achieve the deep decoupling and reconstruction of the spatiotemporal information of dynamic scenes. Driven by the embedded decoupling deformation field, the deformation field contains a coarse-grained trajectory prediction branch and a fine-grained integral fitting branch: the coarse-grained branch combines low-frequency time coding with latent embedding features, enabling Gaussian primitives to capture global rigid motion trajectories and serving as a temporal regularizer; the fine-grained branch uses high-frequency time coding to predict non-rigid deformation and physically fits the motion blur stripes generated by temporal integration through anisotropic scale stretching. An adaptive density control strategy compatible with physics is introduced. Using the physical integral imaging simulation module, high-fidelity and multi-view consistent high-speed dynamic three-dimensional scene reconstruction results are obtained.

The present application relates to a kind of single-pixel imaging method and device of supervised and unsupervised learning combination, belong to the field of computational photography technology.The method designs a kind of single-pixel one-step imaging network based on Transformer, can directly from undersampled single-pixel measurement value reconstruct target image, without the step of solving approximate result, while using the way of supervised learning and unsupervised learning combination to train the network, guarantee the strong generalization of method.Through the combination of mean square loss function and total variation loss function, prevent network from appearing overfitting problem, guarantee the reconstruction performance of method.The device includes data acquisition submodule, supervised learning submodule, unsupervised learning submodule and reconstruction submodule.The present application can effectively capture the long-range dependence between single-pixel measurement value, can realize high-quality single-pixel imaging under low sampling.

The invention relates to the technical field of selfie image optimization, and provides a mobile phone intelligent selfie optimization method comprising the following steps: obtaining original image data of a selfie scene and carrying out feature extraction to obtain a feature map; obtaining corrected key point coordinates according to the feature map; extracting skin color distribution information and illumination intensity information to obtain balanced skin color and illumination parameters; acquiring an ambient light change trend, and generating an illumination compensation coefficient; real-time face tracking and neural network feature extraction are combined to obtain an optimized version of an illumination compensation coefficient; fusing the adjusted image data to obtain a fused image with enhanced details; and extracting and matching the face feature vectors, and outputting an optimized selfie image according to a matching result. Through the cooperation of deep learning and computational photography, the definition, naturalness and individuation effect of the selfie image in a complex light environment are remarkably improved, and the high-quality selfie requirement is met.

The application discloses an inverse ISP method based on a spiral diffusion model and camera perception adaptation, and relates to the technical field of computer photography and image signal processing. The method comprises the following steps: S1, constructing an inverse ISP training sample, each training sample comprising an RGB image, a target RAW image corresponding to the RGB image and a camera label; S2, constructing an inverse ISP network based on a spiral diffusion model, introducing a time-varying weight map related to pixel intensity at different time steps in the diffusion process; S3, setting a camera perception low-rank adaptation module in the inverse ISP network, comprising a plurality of low-rank adaptation branches corresponding to different camera labels, and selecting a corresponding low-rank adaptation branch to participate in network calculation according to the input camera label; S4, in the training stage, training the inverse ISP network based on the forward probability distribution of the spiral diffusion model; in the sampling stage, inputting the RGB image and the camera label into the trained inverse ISP network, and obtaining the target RAW image based on the reverse iterative sampling process of the spiral diffusion model.

A learning device (20) acquires computational photography information about a computational photography camera (101) that captures an image with blur, acquires a normal image captured by a normal camera that captures an image without blur or with little blur and a correct answer label given to the normal image, generates an image with blur based on the computational photography information and the normal image, and creates an image recognition model for recognizing an image captured by the computational photography camera (101) by performing machine learning using the image with blur and the correct answer label.

The invention relates to the technical field of computer vision and computational photography, in particular to a structured scanning platform for quickly capturing a design manuscript and a deviation rectifying method, which comprises the following modules: a multi-modal sensing module for projecting a phase shift stripe sequence and acquiring a modulation image, solving an absolute phase by using a multi-frequency heterodyne method, and obtaining physical point cloud data of the surface of the design draft based on triangulation; and the discrete geometric modeling module constructs the physical point cloud data into a pure complex grid by using Delauni triangulation, calculates the Euclidean distance of the side length of the grid as a discrete Riemannian metric value, and constructs a discrete Hough star operator. In the invention, through an active optical flow closed-loop feedback mechanism, the Lie derivative flow is driven by phase residual error to iteratively correct the discrete Riemannian metric, and then high-fidelity physical expansion is realized, so that the problems that open-loop geometric projection is mostly adopted in the traditional correction technology, and physical precision verification is lacked, so that the correction precision is poor are solved. And therefore, the problems of scale distortion and texture distortion after the precision drawing is digitalized can be solved.

To provide a method and apparatus for computational photography under low light conditions for an image capture device on a mobile computing device.SOLUTION: The method includes receiving, by a mobile computing device, sensor data regarding ambient conditions of a scene during low light conditions of the scene, selecting, based on the received sensor data regarding the ambient conditions of the scene, to capture, using one or more image capture devices of the mobile computing device, a plurality of images of the scene without using a flash, in response to capturing the plurality of images of the scene without using a flash, generating a computed image using the plurality of images of the scene, and providing the generated computed image.SELECTED DRAWING: Figure 8

This invention discloses an inverse ISP image reconstruction method based on a diffusion model, belonging to the fields of computer vision and computational photography. This method uses pure noise as the initial input and introduces the target RGB image as a conditional constraint during the diffusion-based inverse denoising process. By fine-tuning the pre-trained diffusion model using ControlNet, the model learns the mapping relationship between RGB and RAW, thereby gradually guiding the reconstruction of a RAW domain image consistent with the conditional RGB content. To improve the structural consistency, detail fidelity, and color / brightness response of the reconstruction results, this invention incorporates attention-guided strategies (such as spatial attention and cross-feature interactive attention) during the conditional injection process. This fully utilizes the edge, texture, and semantic information in the RGB image, suppressing artifacts and detail loss caused by inversion / reconstruction, and enhancing the ability to restore details in extreme exposure areas such as highlights and shadows.