36 results about "Consecutive frame" patented technology

In multimedia object tracking and merging of tracked objects, an object is tracked through frames of multimedia content until a frame appears in which the tracked object is not detected. A first track is designated as one or more consecutive frames in which the tracked object is detected, the first track ending at the first frame. Tracking continues to try to detect the tracked object in a second frame subsequent to the first frame. If the tracked object is not again detected, information about the first track is output. If the tracked object is detected subsequently, a second track of consecutive tracked object detection is designated. The tracked objects in the two tracks are then compared with the aid of trained data models, and a matching score is determined to reflect the degree of match. If the matching score meets or exceeds a first threshold, the compared tracks are merged using the same identifier assigned to both tracks. If the matching score does not exceed a second threshold that is less than the first threshold, the tracks may be discarded as showing no match. If the matching score falls between the first and second thresholds, an indication is output for further analysis of the compared tracked objects.

The invention provides a continuous frame parameter adjustment method, electronic equipment and a system in on-board diagnostic communication. The method comprises the following steps: determining the priority of UDS service; determining a reference value of the first parameter according to the priority of the UDS service; determining a predicted value of a network load at a future moment through a network load prediction model based on the correlation characteristics of the vehicle at the current moment; and correcting the reference value of the first parameter according to the predicted value of the network load at the future moment and the true value of the network load at the current moment to obtain a corrected value of the first parameter. Through the technical scheme of the invention, the continuous frame parameter can adapt to the dynamically fluctuating network load, the loss probability of the continuous frame is reduced, and the stable and effective transmission of the continuous frame is ensured.

The invention provides a touch action recognition method and device, a storage medium and electronic equipment, and belongs to the technical field of human-computer interaction. The method comprises the following steps: acquiring multi-frame queue data; calculating an effective contact number of the current frame data in the multi-frame queue data, and determining an action type to which the touch action of the current frame belongs based on the effective contact number; if the action type is a point contact type, calculating a corresponding dispersion degree according to the current frame data and the previous frame data, and determining whether the touch action of the current frame is one of a light touch action or a sliding action based on the dispersion degree; if the action type is a beating type or a palm type, calculating the number of continuous frames which are judged to be the same action type in the multi-frame queue data; and determining whether the touch action of the corresponding frame is one of a slow shooting action, a snapshot action, a long press action, a continuous touch action or a single touch action based on the number of the continuous frames. According to the invention, the recognition accuracy of complex gestures can be improved.

A method for compressing a video stream includes retrieving a plurality of frames corresponding to the video stream. For each of two or more sequential frames of the plurality of frames of the video stream, the method includes extracting Key Point Descriptors (KPDs) for the respective frame and processing the respective frame using Principle Component Analysis (PCA) followed by vector quantization, resulting in a quantized explained variance matrix for the respective frame. The quantized explained variance matrix for the respective frame is stored. The KPDs for the respective frame are stored.

A sequence of frames is encoded in the following manner. An evaluation on a frame-by-frame basis is made for at least a zone in a series of successive frames whether the zone should be encoded in an intra-mode or in an inter-mode. A refresh scheme is applied that frequently causes the zone to be forcibly encoded in the intra-mode for one out of several successive frames, irrespective of the evaluation made. The zone in the series of successive frames is quantized so that if, for a frame, the zone is forcibly encoded in the intra-mode and contrary to the evaluation that was made, the zone is quantized with a higher resolution than if the zone were encoded in accordance with the evaluation that was made in one of the following modes: the intra-mode and the inter-mode.

Approaches are disclosed for enhancing frame rate and visual smoothness in real-time video streams through multi-frame interpolation. A classification neural network analyzes two sequential frames, outputting confidence scores that indicate the reliability of motion data for each pixel. These scores determine whether a pixel's motion is accurately described by motion vectors or should be treated as static. The classification results are reused to generate intermediate frames by warping the original frames based on the motion characteristics. Blending weights are calculated by combining warped motion vector confidence values with static values, and a second neural network refines the alignment and blending of candidate frames. This second network predicts intermediate flows and generates new blending weights, which are used to warp and blend the candidate frames, ultimately producing a final interpolated frame that enhances visual smoothness and consistency in the video stream.

The present invention provides a method and device for removing speed blurred false targets, the method comprising: obtaining temporary tracks of continuous Q frames and target point tracks associated with the temporary tracks in each frame, where Q is a positive integer and Q≥2; calculating the distance R between the target point tracks. (i，j) , i and j represent the frame number, i, j∈{1, 2, 3, ..., Q}, i<j; based on the distance R (i，j) and the distance R (i，j) The interval time T (i，j) Calculate multiple distance change rates V (i，j) For multiple distance change rates V (i，j) Data fusion is performed to obtain the true range change rate corresponding to the temporary track; the true range change rate is compared with the defuzzification speed parameter to obtain a speed difference; the speed difference is compared with a preset difference threshold parameter, and based on the comparison result, it is determined whether to upgrade the temporary track to a mature track. The present invention can identify and eliminate false targets with defuzzification errors, avoiding false target output and erroneous upgrade of temporary tracks.

A method for compressing a video stream includes retrieving a plurality of frames corresponding to the video stream. For each of two or more sequential frames of the plurality of frames of the video stream, the method includes extracting Key Point Descriptors (KPDs) for the respective frame and processing the respective frame using Principle Component Analysis (PCA) followed by vector quantization, resulting in a quantized explained variance matrix for the respective frame. The quantized explained variance matrix for the respective frame is stored. The KPDs for the respective frame are stored.

This invention relates to a method for tracking moving ship targets in remote sensing images, applied to remote sensing scenarios with a large field of view and low frame rate. The method compensates for the limitations of single-frame detection features by combining spatial information of the marine ship target with local details of its motion and multi-frame correlation. The process includes continuously acquiring multiple frames of remote sensing images and selecting ship targets to be tracked; determining the identity of the selected ship targets in two consecutive frames; and generating the trajectory of the moving ship target based on its position in the multi-frame images. This achieves high detection performance with relatively low complexity.

Methods and systems of frame based image segmentation are provided. For example, a method for feature object tracking between frames of video data is provided. The method comprises receiving a first frame of video data, extracting a mask feature for each of one or more objects of the first frame, adjusting the first frame by applying each initial mask and corresponding identification to a respective object of the first frame, and outputting the adjusted first frame. The method further comprises tracking the one or more objects in one or more consecutive frames. The tracking comprises extracting a masked feature for each of one or more objects in the consecutive frame, adjusting the consecutive frame by applying each initial mask and corresponding identification for the consecutive frame to the respective object of the one or more objects of the consecutive frame, and outputting the adjusted consecutive frame.

This invention provides a system and method for providing video semantic segmentation. Semantic segmentation is performed on a first frame in a sequence of video frames to obtain at least one first semantic feature of the first frame. Semantic segmentation is performed on a second frame in the sequence to obtain at least one second semantic feature of the second frame, wherein the second frame follows the first frame. Semantic segmentation is performed on a third frame in the sequence to obtain at least one third semantic feature, wherein the third frame follows and is also following the first frame by a first predetermined number of consecutive frames. At least one first semantic feature, at least one second semantic feature, and at least one third semantic feature are combined to form at least one fourth semantic feature of the second frame.

The application relates to a video translation method and device, which comprises the following steps: inputting image features and coordinate features of original data into a fusion input generator; calculating a three-frame loss according to continuous frames of a video of the input generator, and optimizing the generator according to the three-frame loss; and performing video translation based on the optimized generator, so as to improve the visual effect of the converted video.

Disclosed are an encoder for encoding a video sequence having successive frames into a bitstream, a method for encoding a video sequence having successive frames into a bitstream, a decoder for decoding the bitstream to reconstruct the video sequence having successive frames, and a method for decoding the bitstream to reconstruct the video sequence having successive frames.

To generate communication specifications of a device whose communication specifications are unknown.SOLUTION: A communication specification generation device (101) includes a frame extraction unit (11) that extracts a frame having a specific frame length from communication log data generated by a target device (103), a header identification unit (12) that sequentially compares values in a range determined by a standard header length from a head of the frame between two consecutive frames to identify a header of the frame, and a combination identification unit (13) that extracts codes of a command and a response from the header on the basis of position information and size information about the respective codes of the command and response frames and identifies a combination of the codes of the command and the response according to a communication state of the command and response frames in the communication log data.SELECTED DRAWING: Figure 1

The invention discloses a detection method and device, electronic equipment, a storage medium and a computer program product. The method comprises: determining M segments of frame sequences from a video to be detected, M being an integer greater than or equal to 1, each segment of frame sequence in the M segments of frame sequences including N continuous frames, the N continuous frames included in each segment of frame sequence being completely different, and N being an integer greater than 1; for each section of frame sequence in the M sections of frame sequences, performing mask processing on N continuous frames contained in the frame sequence in the to-be-detected video to obtain a mask video; by using data contained in the mask video, sequentially regenerating N continuous frames to be masked according to a sequence from two ends to the middle; detecting the frame sequence by using N continuous frames contained in the frame sequence and the regenerated N continuous frames to obtain a detection result corresponding to the frame sequence; and determining whether the to-be-detected video belongs to a forged video or not by using all the detection results corresponding to the M segments of frame sequences to obtain a first detection result.

The invention relates to a flying dust detection method and device, electronic equipment and a readable storage medium. The method comprises the following steps: respectively acquiring a radar point cloud and a depth map of a current frame through a radar and a depth camera; wherein the radar point cloud and the depth map of the same frame correspond to each other and form a parameter group; updating a frame queue of the current frame based on the parameter group, the frame queue being formed by a plurality of parameter groups corresponding to continuous frames, and the frame queue comprising the parameter group of the current frame; and determining a fusion score of the parameter group of the current frame, and determining a target detection result of the current frame based on the fusion scores of all the parameter groups in the frame queue. By adopting the method, the dust raising environment can be accurately detected.

This disclosure relates to multi-frame interpolation for real-time video processing using deep neural networks. A method is disclosed to improve frame rate and visual smoothness in real-time video streams through multi-frame interpolation. A classification neural network analyzes two consecutive frames and outputs confidence scores indicating the reliability of motion data for each pixel. These scores determine whether the motion of a pixel is accurately described by a motion vector or should be considered static. The classification results are reused to generate intermediate frames by warping the original frames based on motion features. Blending weights are calculated by combining the warped motion vector confidence values ​​with static values, and a second neural network refines the alignment and blending of candidate frames. This second network predicts the intermediate stream and generates new blending weights, which are used to warp and blend candidate frames, ultimately producing final interpolated frames that enhance the visual smoothness and consistency of the video stream.

Techniques for frame-based noise reduction and other signal processing techniques are provided that are able to reduce latency below the frame length. This may be accomplished by using an asymmetric trapezoidal (or quasi-trapezoidal) filter. Instead of cross-fading between successive frames taking up half the frame length, the cross-fading between successive frames is reduced to 10-25% of the frame length, for example. This allows the latency to be reduced to 60-75% of the frame length (plus the signal processing time) instead of the full frame length (plus the signal processing time) in prior techniques.

The present technology is related to an imaging device, an imaging method, and a program capable of imaging in a desired exposure time even when power consumption is reduced. The present invention is provided with: a frame control unit that controls any one of a first frame on which a read operation and a shutter operation are performed, a second frame on which a read operation is performed, a third frame on which a shutter operation is performed, or fourth frames on which neither a read operation nor a shutter operation is performed; and a frame setting unit that sets the third frame as one frame among the fourth consecutive frames. The present technology may be applied to an imaging device, for example.

The invention discloses a multi-frame particle image denoising method based on spatio-temporal feature interaction. The method comprises the following steps: S1, constructing an enhanced data set comprising continuous multi-frame enhanced particle images; s2, constructing a spatio-temporal feature interaction network model, and training the spatio-temporal feature interaction network model based on the enhanced data set to obtain a trained spatio-temporal feature interaction network model; the spatio-temporal feature interaction network model comprises a spatial denoising module and a spatio-temporal denoising module; and S3, based on the trained spatio-temporal feature interaction network model, denoising processing is carried out on actual continuous multi-frame noise-containing particle images. According to the invention, parallel processing is carried out through the space denoising module, then stacking is carried out in the time dimension, denoising enhancement is carried out through the space-time denoising module, continuous multi-frame enhanced particle images are output, the calculation efficiency is improved, and thus high-quality images are provided for subsequent PIV estimation.

A method for identifying stationary regions in frames of a video sequence comprises receiving an encoded version of the video sequence, wherein the encoded version of the video sequence includes an intra-coded frame followed by a plurality of inter-coded frames; reading coding-mode information in the inter-coded frames of the encoded version of the video sequence, wherein the coding-mode information is indicative of blocks of pixels in the inter-coded frames being skip-coded; finding, using the read coding-mode information, one or more blocks of pixels that each was skip-coded in a respective plurality of consecutive frames in the encoded version of the video sequence; and designating each found block of pixels as a stationary region in the respective plurality of consecutive frames.

The invention discloses a bird identification method, which comprises the following steps: according to a bird detection result of a to-be-identified data frame, a bird identification result and continuous frame information of at least one continuous frame of the to-be-identified data frame, at least one effective frame is obtained, and the effective frame is used for representing a data frame which detects a bird target and is stable; under the condition that the frame number of the obtained effective frames is larger than the set frame number, the obtained effective frames are searched according to the set frame number, the bird recognition results corresponding to the currently searched effective frames are counted, the statistical result of the number of times of all the bird recognition results is obtained, and the bird recognition results are used for representing the types of birds; and selecting a bird recognition result with the maximum number of times from the statistical results, and taking the selected bird recognition result as a current recognition result under the condition that the maximum number of times of the selected bird recognition result is greater than a set first number of times threshold value. According to the invention, false detection and false identification can be prevented, and the bird identification precision is improved.

Systems and methods are provided for one or more processors; and memory storing instructions that, when executed by the one or more processors, cause the system to perform: receiving successive frames of sensor data, the successive frames comprising a first frame and a second frame; determining transformations, in sensor coordinates, between coordinates of corresponding elements in the successive frames; determining a mapping between the transformations in sensor coordinates and transformations in geospatial coordinates of the corresponding elements in the successive frames; and determining second geospatial coordinates of the corresponding elements of a third frame based on: a transformation between the second frame and the third frame, and the mapping.

The invention relates to the field of multi-frame recognition, in particular to a safety helmet recognition method based on multiple frames. The method comprises the following steps: S1, collecting a video stream through a camera, and obtaining multi-frame images of wearing and non-wearing of the safety helmet from the video stream; s2, training a detection model by using a YOLOv5 network and training a safety helmet classification model by using VGG; s3, loading the trained detection and classification model by using a deep learning target detection framework TensorFlow; s4, inputting a scene image needing to be detected into a deep learning target detection framework TensorFlow to obtain positions corresponding to a human frame and a head frame; s5, performing preliminary classification on the head frame by using the safety helmet classification model; s6, performing matching tracking on the head frame by using a YOLO detection model; s7, counting the proportion of each tracking target in multiple frames or counting the number of continuous frames existing in the target, and giving an identification result if a certain threshold value is exceeded; through the YOLOv5 network training detection model and the VGG training safety helmet classification model, the accuracy of safety helmet detection in a complex environment is improved.

Systems and methods for universal stage identification for intra-operative and post-operative applications are provided. The system receives a video stream that captures a program of the robotic medical system over a time interval. A system generates a set of consecutive frames from a video stream. The first set of consecutive frames may include a first temporal resolution and the second set of consecutive frames may include a second temporal resolution. The system determines, on a transient basis, a stage of the program within a time interval via a set of consecutive frames input into a first model trained with machine learning. The system inputs stages of the program into the second model to generate stage segments within the time interval. The second model may be trained with machine learning based on historical workflows. The system provides actions based on the metrics of the stage segments.

The present invention discloses a product positioning method, device, and storage medium based on dynamic vision, comprising: obtaining three consecutive frames of images from a video, using a three-frame difference algorithm to obtain an intersecting image of the three consecutive frames; performing dilated convolution processing on the intersecting image to obtain a preprocessed image; shielding areas in the preprocessed image that meet preset conditions, and performing contour extraction to obtain a connected domain; if the width or height of the connected domain is greater than a preset value, intercepting a square area with a maximum side length around the center of the connected domain as a moving product area; if the width or height of the connected domain is not greater than a preset value, intercepting a square area with a side length of the preset value around the center of the connected domain as a moving product area. The embodiment of the present invention uses a three-frame difference algorithm to obtain the intersecting images of the images, performs preprocessing to obtain the preprocessed image, and shields areas in the preprocessed image that meet the preset conditions, which can effectively improve the accuracy of product positioning.

Disclosed is an encoder for encoding a video sequence comprising consecutive frames into a bit stream, a method for encoding a video sequence comprising consecutive frames into a bit stream, a decoder for decoding a bit stream in order to reconstruct a video sequence comprising consecutive frames and a method for decoding a bit stream in order to reconstruct a video sequence comprising consecutive frames.

A method includes determining an operating mode of a first terminal of the one or more terminals; selecting, based on the operating mode, a first frame configuration from a plurality of frame configurations for the first terminal, each of the plurality of frame configurations configuring one or more of: an offset parameter determining the start of a frame, a Channel Occupancy Time “COT” duration in a frame, an idle duration in a frame, a frame duration, a frame period and a gap duration between two subsequent frames; and communicating between the first terminal and the base station and via the first frequency band, using contention-based access and based on the first frame configuration.

This invention belongs to the field of video coding technology, specifically relating to a method for fast coding unit partitioning in 360-degree video based on temporal correlation. The method includes: acquiring and preprocessing cross-frame correlation data, including the current frame CU, the previous frame CU, the luminance difference map between the current and previous frames, the partitioning result of the previous frame, and the partitioning result label; processing the luminance difference map using a similarity network to obtain the similarity between the current and previous frames; selecting, based on the similarity, to directly obtain the coding unit partitioning result from the previous frame's partitioning result or selecting to process the cross-frame correlation data using a multi-level adaptive differentiation prediction network to obtain the coding unit partitioning prediction probability; and determining the partitioning mode based on the coding unit partitioning prediction probability. This invention can operate stably in complex video scenes and under different dynamic scenarios, accurately identify the structural similarity between consecutive frames, and directly reuse the partitioning of the previous frame in highly similar regions, avoiding redundant calculations and improving coding efficiency and stability.

To acquire a plurality of proper samples to be used for prototype-kNN.SOLUTION: An information processing unit 100 calculates, for each pair of two frames 111 succeeding with time, a first index value indicative of how much regions including objects, specified respectively from the two frames 111 of the pair, overlap. The information processing unit 100 calculates, for each pair of two frames 111 succeeding with time, a second index value indicative of how much predetermined kinds of feature quantities of two frames 111 of the pair overlap. The information processing unit 100 specifies the remaining frames 111 by removing a certain number of frames 111 before and after the pairs in which at least first index values are less than a first threshold or second indexes are less than a second threshold from the series of frames 111. The information processing unit 100 extracts frames 111 such that similarity degrees of predetermined kinds of feature quantities to other frames 111 among the remaining frames 111 meet predetermined conditions.SELECTED DRAWING: Figure 1