18 results about "Still frame" patented technology

The present application relates to the field of videos, and specifically relates to a video denoising method and apparatus, and an electronic device and a computer-readable storage medium. The video denoising method comprises: acquiring the number of target static frames of each pixel point in the current video frame, wherein the number of target static frames is the number of frames which are in a video to be denoised and are kept consecutive and relatively static to the current video frame; on the basis of the number of target static frames, constructing a fusion weight corresponding to each pixel point; and on the basis of the fusion weight, fusing the current video frame and the previous video frame, so as to obtain a denoised video frame corresponding to the current video frame. Compared with the prior art, the video denoising method and apparatus, and the electronic device and the computer-readable storage medium provided in the embodiments of the present application have the advantage of improving the video denoising effect.

The invention relates to a display device and a feature value extraction method for static frame detection, and the display device comprises the steps: taking a pixel flow of a target region as an input pixel flow of a current frame image if the target region in the current frame image is a natural region, receiving the input pixel flow of the current frame image, and extracting a feature value of the current frame image; and extracting a pixel characteristic value of the row unit region in the current frame image according to the row unit region, and comparing the pixel characteristic value of the row unit region in the current frame image with a pixel characteristic value of a corresponding row unit region in a previous frame image adjacent to the current frame image to obtain a comparison result so as to determine whether to output a static signal or not. Visibly, the pixel feature value of the corresponding row unit region of the current frame image is extracted according to the row unit region, and the pixel feature value is compared with the pixel feature value of the corresponding row unit region of the adjacent previous frame to obtain the comparison result, so that the difference state of the corresponding regions between the adjacent frames can be accurately reflected; and the accuracy of static frame detection of the display equipment is effectively improved.

The application discloses a still frame detection method, an electronic device and a storage medium. The still frame detection method comprises the following steps: acquiring a to-be-detected image and a first reference image located in a previous frame of the to-be-detected image; if a first moving target is detected according to an inter-frame difference algorithm, the to-be-detected image and the first reference image, determining whether the first moving target is in a typing state according to a first moving direction of the first moving target; if it is determined that the first moving target is not in the typing state, detecting a second moving target according to an optical flow algorithm, the to-be-detected image and the first reference image; and if no second moving target is detected in the to-be-detected image, determining that a current display picture is a still frame picture. The application solves the technical problem of low accuracy of the prior art still frame detection.

The invention relates to a display device and a video static frame detection method, and the display device comprises a controller which is configured to respond to a feature configuration instruction of a user side, and determine at least one target feature algorithm from a plurality of preset feature algorithms; acquiring a pixel block of a video frame to be detected; wherein the video frame to be detected is divided into a plurality of pixel blocks in rows or columns; determining at least one characteristic value of a pixel block in the to-be-detected video frame according to at least one target characteristic algorithm; and comparing the at least one characteristic value of each pixel block in the to-be-detected video frame with the at least one characteristic value of each pixel block in the previous video frame to determine whether each pixel block in the to-be-detected video frame is a static frame picture.

The invention provides a valve opening and closing operation detection method and an applicable system and a readable medium thereof, and relates to the technical field of computer visual recognition. The detection method comprises the following steps: acquiring a plurality of continuous image frames containing the switch component in a preset time period; frame difference values corresponding to a plurality of continuous moments associated with the plurality of continuous image frames are obtained through calculation according to the pixels of the plurality of continuous image frames; a plurality of image frames before the valve is opened and closed for the first time are selected from the plurality of continuous image frames to serve as static frames; obtaining a starting time threshold value and an ending time threshold value corresponding to the same opening and closing operation according to the frame difference values corresponding to the plurality of static frames; and for a plurality of continuous moments corresponding to each opening and closing operation, taking the continuous moment corresponding to the first frame difference value greater than the starting moment threshold value as the starting moment of the same opening and closing operation, and taking the continuous moment corresponding to the first frame difference value less than the ending moment threshold value after the starting moment as the ending moment of the same opening and closing operation.

This application belongs to the field of video and specifically relates to a video noise reduction method, apparatus, electronic device, and computer-readable medium. The video noise reduction method includes the steps of: obtaining the target still frame number for each pixel point in the current video frame; constructing a fusion weight corresponding to each of the pixel points based on the target still frame number; and fusing the current video frame with the previous frame based on the fusion weight to obtain a noise-reduced video frame corresponding to the current video frame, wherein the target still frame number is the number of frames that are continuously and relatively still with respect to the current video frame in the video to be noise-reduced. Compared to the prior art, the video noise reduction method, apparatus, electronic device, and computer-readable medium according to the embodiments of this application have the advantage of improving the effect of video noise reduction.

Method in a data processing system (200) comprising at least one processor and at least one main memory, wherein the at least one main memory contains instructions that are executed by the at least one processor to configure the at least one processor to implement a mechanism (302) for detecting action-object interactions in order to detect actions in overloaded video scenes from videos in the real world, the method comprising: in response to receiving an action-object interaction video (304) with a corresponding label (308): computing an object frame around an object of interest (402) named in the corresponding label in a start frame of a set of still frames (306a-306n) in the action-object interaction video where the object of interest occurs in the still frame (502-516);Forwarding (518) the object frame from the start frame to a subsequent frame of the set of frames (306a-306n); for the start frame and the subsequent frame with an object frame in the action-object interaction video (304): refining (520) the object frames of the start frame and the subsequent frame using knowledge of the object of interest (402) to form associated refined object frames; and cropping (522) the start frame and the subsequent frame based on the associated refined object frames to form a set of cropped frames;Processing (524) the set of cropped frames in the same order as in the action-object interaction video (304) to determine a probability that an action to be tested is performed in accordance with the corresponding label (308), wherein the action to be tested includes opening and / or closing a filing cabinet; and in response to a determination that the probability of the action to be tested in accordance with the corresponding label (308) is equal to or exceeds a test threshold: providing confirmation (528) that the action-object interaction video (304) is performing the action to be tested.

An image capture system comprises: (a) a camera; (b) a camera mounting, configured to provide the camera with at least two angular degrees of freedom; (c) a mount base, configured to enable a base of the camera mounting to be stationary during performance of imaging; and (d) a movement controller, operatively coupled to the camera mounting and configured to repeatedly instruct the mounting to move the camera to a next imaging position. These instructions comprise a first instruction comprising movement in at least a first angular DOF and a second instruction comprising movement in at least a second angular DOF. These move the camera to orientations which enable a capture of a 2-D set of still image frames of a body surface. Still frames, and at least some corresponding adjacent frames, have an overlap amount of more than 50% of an image frame dimension.

The application discloses a still frame detection method and device, comprising: obtaining an Nth image group G N ; decoding a key frame in the obtained image group G N ; determining the similarity between the key frame in the image group G N and the key frame of an (N-1)th image group G N‑1 ; and determining whether there is a still frame based on the similarity and whether there is a still frame starting moment, wherein the scheme is realized based on a skip frame detection strategy of the key frame, that is, the still frame detection is preferentially performed on the key frame in the video signal, and if no still frame is found, the detection of other frame types in the current image group is directly skipped; the implementation not only reduces the overhead of the decoder in processing the non-key frame, but also greatly reduces the frame data required for the still frame detection, so that the detection speed is greatly improved; and the efficient detection manner significantly reduces the occupation of the server resources, and reduces the hardware cost and energy consumption.

The invention provides a fire hazard risk adaptive monitoring method based on YOLOv12-AFE, and relates to the technical field of fire hazard monitoring, and the method comprises the following steps: collecting a video stream, carrying out the preprocessing of data, and carrying out the marking of the data; rejecting static frames by adopting a lightweight background modeling method or a motion detection method; a YOLOv12-AFE model is constructed, the YOLOv12-AFE model comprises a backbone network in which an adaptive feature enhancement module AFE is embedded, and the AFE module comprises a convolution embedding module, a space context module, a feature refining module and a convolution multilayer perceptron; the YOLOv12-AFE model is trained through an Inner-IoU loss function, and bounding box regression is optimized; and the fire risk is monitored by using the trained YOLOv12-AFE model. According to the technical scheme of the invention, the problem of weak feature extraction capability for cluttered backgrounds, small targets and semitransparent objects in the prior art is solved.

The invention relates to the field of videos, in particular to a video noise reduction method and device, electronic equipment and a computer readable storage medium. The video noise reduction method comprises the steps that the target static frame number of each pixel point in a current video frame is acquired, and the target static frame number is the frame number which is kept continuous with the current video frame and is relatively static with the current video frame in a video to be subjected to noise reduction; constructing a fusion weight corresponding to each pixel point according to the target static frame number; and fusing the current video frame and the previous video frame according to the fusion weight to obtain a noise reduction video frame corresponding to the current video frame. Compared with the prior art, the video noise reduction method and device, the electronic equipment and the computer readable storage medium provided by the embodiment of the invention have the advantage that the video noise reduction effect can be improved.

The present application relates to the field of videos, and specifically relates to a video denoising method and apparatus, and an electronic device and a computer-readable storage medium. The video denoising method comprises: acquiring the number of target static frames of each pixel point in the current video frame, wherein the number of target static frames is the number of frames which are in a video to be denoised and are kept consecutive and relatively static to the current video frame; on the basis of the number of target static frames, constructing a fusion weight corresponding to each pixel point; and on the basis of the fusion weight, fusing the current video frame and the previous video frame, so as to obtain a denoised video frame corresponding to the current video frame. Compared with the prior art, the video denoising method and apparatus, and the electronic device and the computer-readable storage medium provided in the embodiments of the present application have the advantage of improving the video denoising effect.

The invention provides a digital human video action continuous splicing method and device for a streaming live broadcast scene, and relates to the technical field of virtual digital humans, and the method comprises the steps: obtaining a digital human video in which a digital human is still and does actions alternately, and constructing a frame track; when the person state is a static state, each frame is extracted and directly added to the tail part of the playing queue, and when the person state is a speaking state, after extracted continuous video frame segments are processed by a video frame processing module, the processed continuous video frame segments are sequentially added to the tail part of the playing queue; when the person state is a static state, the extracted video frames are all static frames, and when the person state is a speaking state, the last frame in all the extracted video frames is a static frame; and displaying the content of the playing queue frame by frame, and synchronously playing the original audio. According to the invention, the picture jitter during the action switching of the digital person can be reduced, and the stability and natural expressive force of the digital person in the streaming live broadcast are improved.

Image decoding device and method thereof. The embodiments of the present application provide a still frame image decoding method, which comprises: decoding a first image frame in a first data from a plurality of data according to a reference frame in a reference location of a memory, and storing the first image frame in the reference location to cover at least a part of the reference frame; selecting the first data from the data in response to a still frame command, and decoding a second image frame in the first data according to the first image frame in the reference location; adjusting the second image frame, and storing the adjusted second image frame in an output location of the memory; keeping the first image frame in the reference location in response to the still frame command, so as to decode the second image frame again according to the first image frame in the reference location; and decoding the second image frame again according to the first image frame in the reference location in response to an end command, so as to store the second image frame in the reference location to cover at least a part of the first image frame.

The application provides a digital human video action continuous splicing method and device for live streaming scene, and relates to the technical field of virtual digital human, which comprises the following steps: acquiring a digital human video in which a digital human alternately keeps still and performs actions, and constructing a frame track; when the state of the character is a still state, every frame is directly added to the tail of a playing queue; when the state of the character is a speaking state, a continuous video frame segment is processed by a video frame processing module, and then the processed continuous video frame segment is sequentially added to the tail of the playing queue; when the state of the character is a still state, the extracted video frame is a still frame; when the state of the character is a speaking state, the last frame of all the extracted video frames is a still frame; the content of the playing queue is displayed frame by frame, and the original audio is synchronously played. The application can reduce picture jitter when the action of the digital human is switched, and improve the stability and natural performance of the digital human in live streaming.

The invention discloses an edge end fire fighting access occupation detection method, which comprises the following steps: S10, motion vector extraction and static scene identification: inputting a video source into a video encoder by taking a single-frame image as a unit, obtaining motion vector information of all macro blocks of the frame image after each frame image is encoded, and judging whether the frame image is a static frame or not; m continuous static frames form a static scene SN, and the last frame of the static scene SN is marked as a representative frame F1; s20, motion event detection and new static scene extraction: after a motion frame is detected, waiting for continuous M static frames to form a new static scene SN2, and marking the last frame as a representative frame F2; s30, foreground region extraction: respectively extracting foreground masks mask1 and mask2 of F1 and F2 by using a deep learning background removal model; s40, foreground difference comparison and occupation determination: calculating a difference matrix D (x, y) = mask2 (x, y)-mask1 (x, y) for the mask1 and the mask2 pixel by pixel; if D (x, y) = + 1 represents that the foreground is newly added, channel occupation is determined; and if D (x, y) =-1 indicates that the foreground disappears, judging that the channel is unoccupied.

The invention provides a static frame transmission method and device based on Hash comparison, equipment and a medium, and the method comprises the steps: a collection end collects a video frame in real time, and carries out the pixel-level comparison of a current frame and a locally cached historical reference frame; if the two pixels are completely consistent, calculating a hash value of the current frame and only sending a static frame data packet containing the hash value to a receiving end; after receiving, the receiving end compares the hash value with a hash value corresponding to a locally cached historical reference frame; if the comparison is consistent, the receiving end directly and repeatedly outputs the locally cached historical reference frames; and if the Hash comparison is inconsistent or the acquisition end judges that the pixels of the current frame and the historical reference frame are inconsistent, the receiving end receives the complete coded video stream sent by the acquisition end, and outputs and updates the locally cached historical reference frame and the corresponding Hash value after decoding. According to the invention, through combination of pixel-level comparison and Hash verification, the data transmission amount is reduced, the bandwidth is saved, and the transmission efficiency is improved.