18 results about "Temporal filtering" patented technology

In various embodiments, when reference picture resampling is applied to achieve better coding efficiency, methods and devices adapt temporal filtering of a noise-reduced input video prior to encoding according to a resolution ratio. According to one embodiment, temporal filtering strength values are fine-tuned at a pre-processing module according to the resolution ratio of the RPR mode. According to another embodiment, temporal filtering is deferred after rescaling when the RPR mode is enabled, and filtering strength values are fine-tuned according to the resolution ratio of the RPR mode. A third embodiment proposes to refine motion estimation in the encoder by setting the block size of motion estimation in temporal filtering according to the resolution ratio of the RPR mode.

This disclosure provides video encoding methods, apparatus, electronic devices, storage media, and programs. [Solution] The specific solution includes: obtaining a reference frame adjacent to the current frame and in which a hierarchical partitioning of the initial region of interest (ROI) has been established; reusing motion vectors generated during the motion-compensated time filtering process in the reference frame to determine the predicted ROI position of the current frame; extracting sensory feature points from the current frame and correcting the predicted ROI position based on the feature point spatial topology relationship in the initial ROI; assigning differentiated quantization parameters to the corrected primary face region, secondary face region, and background region, respectively; dynamically allocating the bitrate of the three layer regions based on real-time network bandwidth; and outputting the encoded frame of the current frame and associated ROI hierarchical metadata.

The application relates to a time filtering method, device, computer equipment, chip and medium, which comprises the following steps: acquiring a to-be-filtered signal corresponding to a current filtering number; determining target reference data corresponding to the current filtering number according to the current filtering number; determining a time filtering coefficient corresponding to the current filtering number according to the target reference data; and performing filtering processing on the to-be-filtered signal corresponding to the current filtering number according to the time filtering coefficient. The application can improve the filtering effect.

A system and method for identifying container numbers on front-mounted cranes are disclosed, relating to the field of container number identification on front-mounted cranes. To address the shortcomings of existing methods, such as inability to adapt to rainy or snowy weather and low-light conditions, poor container number identification performance, and poor anti-interference capabilities, this invention acquires video or images of containers on front-mounted cranes; extracts continuous frames from the video at a fixed frequency and performs grayscale processing on each frame to reduce computational complexity; applies temporal filtering to the preprocessed video / image; applies spatial filtering to the filtered video / image to remove isolated noise points; reduces high-frequency noise and smooths the image through weighted averaging; enhances image contrast by adjusting the grayscale distribution of the image, making the container number area clearer; and performs container number area detection and character recognition on the processed image to achieve automatic container number identification. This invention is mainly used for identifying container numbers on front-mounted cranes.

This invention provides a method and apparatus for superconducting motor quench detection based on sliding time filtering, relating to the field of superconducting motor quench detection technology. The method includes: differentially processing the terminal voltage signals of adjacent superconducting rotor magnets in the superconducting motor; conducting three trial runs of the superconducting motor before quench detection to obtain interference frequency data; establishing sliding time windows with both long and short time scales during the actual operation of the superconducting motor, and performing notch filtering on the differential voltage signals of adjacent superconducting rotor magnets; performing linear regression fitting on the filtered signals within the sliding time windows to extract effective quench signal feature values; and establishing a quench protection criterion based on the quench feature values. This invention solves the problem of achieving high reliability and rapid response in quench detection technology under the complex interference environment of superconducting motor rotation operation, a problem that is difficult to address in existing technologies.

This paper presents a method for fusing and reconstructing trajectory data from multiple sources in hybrid traffic flow, relating to the fields of multi-source sensor fusion localization and target tracking. The method acquires the 3D acceleration sequence and RGB video stream of the target object, extracting the geometric center coordinate sequence. Based on the frequency domain energy distribution of acceleration, the target's fundamental frequency parameters are obtained. Temporal filtering coefficients are constructed using the dynamic sampling rate of the video, and the geometric center coordinates are filtered to obtain a corrected center coordinate sequence, which is then mapped to a measurement vector. Kinematic state prediction is performed using acceleration to obtain prior inferred state variables, and then posterior updates are performed using the measurement vectors to obtain the optimal spatial position. The reconstructed trajectory sequence is generated by stitching the data together, and the sliding time window is updated based on feedback-optimized parameters. This method effectively eliminates spurious displacements caused by gait, fundamentally eliminating jagged oscillations in the trajectory, resulting in a smooth and stable fused trajectory.

The application discloses a kind of based on motion compensation time domain filtering method for image, comprising: using autoregressive moving average model to the obtained gyroscope error is modeled, the change rule of data is used to estimate model parameter, and gyroscope error model is established;Using the gyroscope data of removing error obtained by gyroscope error model, the rotation matrix of camera is calculated, and the rotation image is compensated in reverse using rotation matrix, and the de-rotation image is obtained.Through kalman filtering to image motion vector sequence, image random jitter is obtained, and the de-rotation image is compensated according to random jitter, and the stable image and the subjective motion vector of image are obtained;Using the motion target detection algorithm in difference map based on large map coordinate system to detect target in stable image.The application has high detection sensitivity, and can detect target and its trajectory quickly through a small amount of frame track correlation.

Provided is a method for encoding a video, an electronic device and a storage medium, relating to the field of video encoding. The method includes: obtaining a reference frame that is adjacent to a current frame and has established hierarchical division of an initial ROI; reusing a motion vector generated by the reference frame in a motion compensated temporal filter process to determine an ROI position prediction value of the current frame; extracting facial feature points from the current frame, and correcting the ROI position prediction value based on a spatial topological relationship of feature points in the initial ROI; configuring differentiated quantization parameters for a main face region, a secondary face region and a background region after correction respectively; dynamically allocating a three-region bitrate according to a real-time network bandwidth; and outputting an encoded frame of the current frame and associated ROI-level metadata.

This invention discloses a method for characterizing deep carbonate reservoirs based on statistical level compression transformation. The method includes: first, inputting the seismic signal to be analyzed, selecting an analysis window function and setting the Taylor expansion order, and constructing a time filter. After calculating the short-time Fourier transform and weighted transform based on the frequency domain window function, convolution filtering is performed using the time filter to construct autocorrelation and cross-correlation statistics. Based on this, a system of linear equations is established to solve for the statistical group delay estimate, and then the short-time Fourier transform coefficients are rearranged in time and frequency to obtain the statistical level compression transformation result. The seismic data is then subjected to time-frequency transformation using the above method to obtain the statistical level compression transformation result. From the time spectra of different seismic traces obtained by this transformation, time series corresponding to constant frequencies are extracted and synthesized into profile maps to characterize reservoir reflection interfaces and delineate reservoir boundaries.

The invention discloses a two-degree-of-freedom helicopter preset performance tracking control method based on reinforcement learning. The method comprises the following steps: constructing a two-degree-of-freedom helicopter dynamic model; a preset performance function is introduced, a constrained tracking error is converted into an unconstrained form through error conversion, and therefore it is guaranteed that the system reaches a steady state within preset time; designing a preset time nonlinear disturbance observer; based on a backstepping control architecture, combining with a predefined time filter, and decomposing the system into two stages of subsystems through coordinate transformation; an optimal performance index function is constructed for each stage of subsystem, an optimal control strategy is approached online by using a reinforcement learning algorithm, and an optimal virtual controller and a self-adaptive optimal actual control law are designed. According to the method, all signals in a two-degree-of-freedom helicopter closed-loop system can be converged and bounded within predefined time, transient and steady state response indexes of the system are improved, and it is ensured that the helicopter can still rapidly and accurately track a reference trajectory under complex working conditions and quantized limited conditions.

A lock-detection circuit designed for integration into a phase-locked loop or a delay-locked loop includes a first signal-duration filter processing an up signal (UP) to generate a filtered UP signal containing segments that exceed a minimum duration, and a second signal-duration filter handling a down signal (DN) to produce a filtered DN signal with segments surpassing the minimum duration. Additionally, the circuit includes first and second timers receiving the filtered signals, and an output logic circuit triggering a LOCK detected signal when both timers are activated. The minimum duration is set to exceed spikes in the UP and DN signals, accounting for jitter, and is longer than the expected duration of a phase error correcting pulse when the loop is in lock.

Embodiments of the present application provide an image denoising method and an image processing device. The method comprises: obtaining a tth input image and a tth reference image; determining a first weight map based on the tth input image and the tth reference image; performing a guided filtering operation on the first weight map with the tth input image as a guide image to generate a second weight map; and performing a temporal filtering operation on the tth input image and the tth reference image based on the second weight map to generate a tth output image.

This invention relates to the fields of artificial intelligence and robot perception technology, specifically a method and system for dynamic calibration of extrinsic parameters in a distributed sensor network based on spatial intelligence. The method extracts atomic spatiotemporal events from each sensor and generates event features to construct a randomly initialized correlation field matrix. The correlation field state corresponding to the current time step, the event feature set, and the time step encoding are input into a pre-trained correlation field denoising network, iteratively obtaining the final correlation field state. High-confidence homologous event counterparts are extracted, three-dimensional point pairs are constructed, and the extrinsic parameter matrix is ​​solved using a random sampling consensus algorithm. The method is repeatedly executed over a sliding time window and combined with temporal filtering to achieve dynamic updates. This method can improve the robustness and accuracy of extrinsic parameter calibration in complex dynamic environments and under non-overlapping field-of-view conditions.

A brain-computer interface (BCI) signal processing method and system are disclosed. The method includes: receiving brain activity detection signals from a target object; mapping a parameter matrix G to an in-memory computing array, wherein a first matrix H, a second matrix W, and a third matrix T are obtained based on a training model including multiple stimulus task templates to perform temporal filtering, spatial filtering, and template matching on the brain activity detection signals; calculating a parameter matrix G, G = TWH, for one-step decoding of the brain activity detection signals; using the in-memory computing array mapping the parameter matrix G to perform one-step decoding of the brain activity detection signals to obtain a decoding result corresponding to the brain activity detection signals; determining, based on the decoding result, a control command for the target stimulus task corresponding to the brain activity detection signals from multiple stimulus task templates; and sending the control command to execute the task. This method reduces the computational load and latency of the algorithm, and decreases circuit area and energy consumption.

This application discloses an ambient light synchronization method, apparatus, and device based on video content color analysis, comprising: obtaining the current video frame and its display timestamp from the video stream; downsampling the video frame; generating a pixel weight mask for eliminating interference areas based on the downsampled video frame; extracting candidate primary colors of the downsampled video frame based on the downsampled video frame, the pixel weight mask, and a preset spatial weight matrix; performing temporal filtering and perceptual mapping on the candidate primary colors to obtain the luminaire-driven color; generating and sending an ambient light synchronization command to at least one target luminaire based on the luminaire-driven color and its target effective time, to instruct the target luminaire to synchronize its emission state to the luminaire-driven color at the target effective time. This application can achieve lightweight computation, interference-resistant primary color extraction, smooth flicker-free operation, color gamut adaptation, and precise time-series alignment for video-linked ambient light synchronization.

This invention provides a spatiotemporal video denoising method based on PID control, comprising: S1. Spatial filtering of the current frame; S2. Spatial filtering of the reference frame; S3. Calculating the spatial filtering weights of the reference frame: calculating the optimal spatial filtering weights of the reference frame using a PID adjustment method; S4. Calculating the temporal filtering weights: calculating the optimal temporal fusion weights of the reference frame using a PID adjustment method; S5. Spatiotemporal fusion to obtain the final processing result. In this method, the PID adjustment method smoothly adjusts the weighted weights of the reference frames in moving and stationary regions, making the transition between the imaging effects in moving and stationary regions smoother and the visual effect more natural. The integral control part of the PID adjustment method can effectively eliminate the problem of inaccurate motion judgment caused by noise, and can effectively avoid the phenomenon of motion blur.