116 results about "Cropping" patented technology

The invention provides an automatic image cutting method and system based on face key points, which are applied to the technical field of image processing, and the method comprises the steps: inputting a portrait image into a portrait segmentation model for foreground extraction, and obtaining a figure region mask; performing top positioning estimation based on the figure area mask to obtain a top position of the portrait image; inputting the portrait image into a face key point detection model to obtain a plurality of key points; determining a face midpoint of the portrait image based on the position relationship of the plurality of key points; performing affine transformation estimation based on the source point set and a preset target point set to obtain an affine transformation matrix; adjusting the vertical offset of the affine transformation matrix to obtain a longitudinal correction matrix; performing edge correction on the longitudinal correction matrix to obtain a cutting transformation matrix; and performing image transformation on the portrait image based on the cutting transformation matrix to obtain a standard composition image. According to the invention, the cut image with uniform size, standard composition and good detail retention can be generated.

The invention belongs to the field of image processing, particularly relates to an automatic association adjustment method for ground identification lines based on block division, and aims to solve the problems of long updating period, potential errors and omission, high resource consumption and low efficiency. The method comprises the following steps: analyzing airport AIP vector data to generate a structured data set; and constructing a topological sub-region in each functional region according to the connection relation of the marked line end points, subdividing the grid units when the marked line density exceeds a threshold value, and outputting a layered region map. And taking the area / grid unit as a detection unit, detecting new and old AIP change areas and position identifiers, and generating new version change area data. Recursively searching an N-degree correlation line by taking the change area as a starting point, and only updating the correlation line of geometric, type or topological connection change. Finally, the changed data and the unchanged data are spliced, and a complete airport model is generated through boundary adaptive extension cutting and global topology reconstruction. According to the method, efficient and accurate updating of the airport ground identifier is realized.

The invention discloses a cross-modal anti-attack method and system based on mask weight and random cutting, and belongs to the technical field of deep learning. The method comprises the following steps: extracting a saliency mask of a source image; disturbing the source image by using the disturbance variable to generate an adversarial sample; randomly sampling K cutting frames in combination with a significance mask; respectively extracting clipping characteristics of the clipping frame on the adversarial sample and target embedding selected on the target image according to an alignment mode; calculating the cutting loss of the cutting frame according to the cutting feature and the target embedding; calculating the weighted sum of the cutting loss according to the significance mask to obtain the total loss; performing back propagation according to the total loss so as to update the disturbance variable; and re-perturbing the source image by using the perturbation variable based on the updated perturbation variable until the maximum step number or total loss convergence is achieved. The method gives consideration to attack success rate, mobility and disturbance concealment, and can adapt to quality evaluation of different large language models.

The invention relates to the technical field of industrial detection, and provides a steel ladle air brick defect detection method and system based on image recognition, and the method comprises the steps: carrying out the preprocessing of a target steel ladle air brick image shot by an industrial camera, including edge feature extraction, cutting and enhancement, and obtaining a preprocessed image set; for the block corresponding to each preprocessed image, acquiring a multi-angle image set with different resolutions shot by a multi-angle sensor; inputting the multi-angle images into a defect feature extraction network to generate a feature image set, and integrating to obtain an integrated feature image; generating defect positioning and type information based on the integrated feature map and the defect identification model set; and finally, combining the initial image, the model set and the defect information to generate a defect detection result. According to the method, the defect detection precision is improved through multi-angle image fusion and model collaborative verification. According to the invention, the accuracy and reliability of steel ladle air brick defect detection can be improved, and the accuracy of defect positioning and classification is enhanced.

A system for recognizing vertically oriented alphanumeric text in images, the system including a processor configured to receive one or more images comprising vertically oriented alphanumeric text and detect one or more regions-of-interest in each image via a trained text detector. The processor is configured to execute a cropping of the detected one or more regions-of-interest encompassing vertically oriented alphanumeric text from each image to obtain one or more text crop portions and rotate the one or more text crop portions to obtain one or more orthogonally rotated text crop portions. The processor is configured to execute a trained ensemble of two different text recognition models on each of the obtained one or more text crop portions and the one or more orthogonally rotated text crop portions and generate a set of candidate recognized text strings based on the executed trained ensemble and determine a final recognized text string.

The invention provides a display device and a picture cutting method. After the display device enters the cutting mode in response to the cutting instruction, when the size of the first original image is larger than the size of the display area, the first original image can be zoomed according to the first zooming proportion to obtain the second image, and the first preview image corresponding to the cutting area is previewed in real time in the cutting process with the second image as the preview basis. When the user confirms cutting, the first preview image can be mapped to the first sub-original image in the first original image based on the first scaling, and the content consistency of the first sub-original image and the first preview image is ensured. And cutting out a first sub-original image from the first original image, and displaying the first sub-original image in the display area after zooming processing according to the first zooming proportion so as to ensure that the display effect of the zoomed first sub-original image is consistent with the display effect of the first preview image.

The application discloses a target detection domain adaptation method based on background and foreground cropping and interchanging. The application adopts a teacher-student model framework of semi-supervised learning, slowly updates through an exponential moving average of a teacher model, enables the model to obtain information from main knowledge learned by a student model, and thus realizes a self-supervision effect. A domain discriminator is introduced to perform adversarial training, enhances the ability of the model to extract domain-invariant representation, and thus improves robustness. In addition, a foreground-background mixed instance strategy further enhances the domain adaptation capability of the model, effectively creates target domain instance images with source domain backgrounds and source domain instance images with target domain backgrounds through cropping, padding and pasting operations. Finally, an iterative optimization strategy is adopted to continuously improve the model performance, a student model is updated through gradient back propagation by calculating multiple loss functions, and a teacher model is updated using an exponential moving average.

The invention provides an electronic image stabilization control method, a program product, electronic equipment and a storage medium, and the method comprises the steps: converting a current frame image based on a homography matrix corresponding to the current frame image, and obtaining a first image region; determining whether a preset cutting area in the current frame image is completely located in the first image area or not; in response to the fact that the preset cutting area is not completely located in the first image area, adjusting the homography matrix; and carrying out conversion again based on the adjusted homography matrix until the preset cutting area is completely located in the first image area, and carrying out cutting based on the preset cutting area to obtain a stable image corresponding to the current frame image.

The application discloses a kind of provincial range rural black and odorous water body satellite remote sensing identification method and system, it is related to environmental remote sensing technical field;A kind of provincial range rural black and odorous water body satellite remote sensing identification method, comprising the following steps: obtaining data, data processing, data image inlaying, obtaining water body distribution data, data cropping, obtaining index data product, obtaining classification grid, preliminary obtaining black and odorous water body distribution, obtaining the final distribution of black and odorous water body;The present application utilizes multi-source satellite remote sensing image data, in combination with water quality parameters, constructs the remote sensing identification model of provincial range rural area water body using NIR, NDWI, SWI index and slope data etc., then uses BOI, NDBWI etc. Index constructs rural area suspected black and odorous water body extraction model, in combination with the relevant data of water body field measurement in monitoring area determines the extraction threshold of water body and black and odorous water body, realizes the automatic identification and extraction of rural area black and odorous water body, and develops relevant software system.

This invention provides an information processing method that can generate an image from an image containing multiple people, with the main person in the composition being the central figure, according to the type of event depicted in the image. [Solution] An information processing method according to one embodiment of the present disclosure acquires an overall image in which multiple people are pictured, identifies the event type of the overall image, identifies a target person from the multiple people according to the identified event type, determines the area containing the target person as a cropping area, and generates individual images by cutting out the cropping area.

The application provides an image dynamic cropping and frame buffer method and device, and relates to the technical field of image processing. The method comprises the following steps: in response to an image display task, acquiring a to-be-processed image frame sequence, cropping requirement information and display adaptation information; determining a buffer write control rule and a cropping read control rule according to the cropping requirement information and the display adaptation information; alternately writing the image frame into a first buffer area and a second buffer area based on the buffer write control rule; when the image frame is written in one buffer area, reading buffer frame data corresponding to a target cropping area from another buffer area in which the image frame has been written to obtain cropping frame data; and performing output adaptation processing on the cropping frame data according to the display adaptation information to obtain display image data. Through the method provided by the application, the cropping area adjustment flexibility can be improved, the image buffer and cropping processing efficiency can be improved, and the adaptability between the cropped image and the display output can be improved.

An image cropping method includes: displaying a to-be-cropped image on a first screen, and receiving a first input performed on a second screen in a case that a crop box is displayed on the to-be-cropped image; in response to the first input and in a case that the relative position of the to-be-cropped image and the crop box is kept unchanged, adjusting the size of the to-be-cropped image and the size of the crop box according to the same contraction or magnification ratio; receiving a second input performed on a target border line of the crop box displayed on the first screen; and in response to the second input and in a case that the display position of the to-be-cropped image is kept unchanged, adjusting the position of the target border line and cropping the to-be-cropped image.

The invention discloses a reconfigurable crop image processing method and system based on software and hardware cooperation, and belongs to the technical field of embedded systems and image processing. The problems that in the prior art, an FPGA neural network accelerator model is solidified, the hardware resource utilization rate is low, and the detection precision and the reasoning speed are difficult to consider at the same time are solved. According to the method, the complexity score of an input image is calculated through an edge detection operator, a color histogram and a gray level co-occurrence matrix, a rapid detection mode, a fine recognition mode or a cooperative reasoning mode is selected according to the complexity score and delay constraint, and dynamic partition configuration is carried out on a processing unit array in the reconfigurable accelerator; in a cooperative reasoning mode, executing the lightweight target detection model through the first processing unit group to quickly detect and output a candidate box, extracting a region-of-interest feature map through the ROI cutting unit, routing the region-of-interest feature map to the second processing unit group, executing the high-precision target detection model to perform fine recognition, and fusing double-model output to generate a final detection result; the two models realize hardware resource sharing through a three-level weight storage architecture. The precision and speed balance capability of agricultural image detection are effectively improved, the resource utilization rate is improved, and the method can be applied to intelligent agricultural scenes such as crop disease recognition and fruit grading.

This invention relates to the field of image cropping, specifically disclosing a method, system, terminal, and medium for segmenting an image to generate a jigsaw puzzle cropping path. The method involves initial segmentation of elements based on element parameters, dividing each element into rectangles; these parameters include the number of rows, columns, and image size. The coordinates of the four corners of each element are obtained from the initial segmentation results and stored in a coordinate file. The concavity / convexity state of each edge of each element is configured. A cropping path is generated based on the concavity / convexity state of each edge of each element and the coordinates of its four corners. This invention enables individual editing of any jigsaw puzzle element, thereby generating a cropping path that can be directly applied to the generated image, improving generation efficiency and reducing costs.

The technology provides for enhanced image cropping via in-context learning. It includes an efficient prompt retrieval mechanism for image cropping to automate the selection of in-context examples. It also includes an iterative refinement strategy to iteratively enhance the predicted crops. The image cropping framework is applicable to a wide range of cropping tasks, including free-form cropping, subject-aware cropping, and aspect ratio-aware cropping. The approach employs a trained large vision-language model associated with in-context learning. For instance, given an input image (whether from free-form, subject-aware or aspect ratio-aware cropping), the top-K semantically similar images from a dataset are retrieved as an in-context learning prompt. Then the in-context learning prompt is fed to a pretrained vision-language model to generate a set of crops. The crop candidates of the set are iteratively refined to yield a final output crop. The final output crop can then be applied to a downstream imaging task.

The application provides a video image fast clipping and dimension reduction method and system based on a hierarchical bounding box, applied to the technical field of image processing, wherein the method collects multi-view dynamic scene point cloud data, and then divides the space by using a hierarchical bounding box to establish a bounding box structure with density characteristics; subsequently, optical diffraction array pre-filtering is used, and point clouds are separated based on wavelength selectivity and density distribution to generate a first subset and a background subset; then, adaptive pulse code modulation is used on nodes in the first subset to obtain high-precision data, and nodes in the second subset are compressed into low-code-rate data by using run-length encoding; subsequently, the two types of data are input into an improved k-d tree based on the spatial relationship of the bounding box to generate a two-dimensional point set representing key features, and a spatial index is constructed to realize real-time scene retrieval. The application can realize image dimension reduction while retaining key geometric features of the image, thereby improving efficient real-time spatial retrieval in a dynamic scene.

The application provides a lane line recognition processing method and device and a vehicle terminal, and relates to vehicle driving technology. The method comprises the following steps: collecting an original image in front of a road; wherein the original image comprises a plurality of lane lines to be recognized; performing scaling processing on the original image to obtain a first image after scaling, inputting the first image into a preset resize model for lane line detection to obtain a first detection image; performing cropping processing on the original image to obtain a second image after cropping, inputting the second image into a preset crop model for lane line detection to obtain a second detection image; and performing fusion recognition processing on the first detection image and the second detection image to obtain a target detection result about the lane lines to be recognized. The method greatly improves the detection accuracy of remote lane lines and solves the technical problem of low detection accuracy of remote lane lines.

The invention provides a redried tobacco strip grading method and system based on computer vision, and belongs to the technical field of tobacco processing and quality control. The method comprises the following steps: firstly, detecting working conditions, including detecting the position of a pressure head, the position of a smoke box, obstacles, the working state of a press and the like; after the working conditions are met, the camera obscura is started to move to the position vertically above the smoke box; starting a light source and a camera, and collecting a clear tobacco strip image; preprocessing operations such as denoising, enhancement, cutting and color correction are carried out on the collected image; abnormal conditions including box covering, exposure, local shielding, leaf breaking and the like are detected; inputting a grading algorithm to the normal condition image, and outputting tobacco lamina grade and matching degree information; and visualizing a tobacco strip analysis result, aligning the result with the image data, and storing the result in a database. According to the method and the system, automation, high efficiency and accuracy of tobacco strip data acquisition and grading are realized, abnormal conditions can be effectively identified and processed, and the method and the system are suitable for online quality control of a tobacco production line.

The invention relates to a coding and decoding device, a coding and decoding method, a bit stream transmission device and a storage medium. This encoding device is provided with: a circuit that, during operation, designs each of a plurality of sub-pictures included in a picture, encodes configuration information indicating the configuration of each of the plurality of sub-pictures included in the picture, and stores the encoded configuration information; the circuit encodes each of the plurality of sub-pictures, the circuit, in action, designs each of the plurality of sub-pictures such that at least one pixel contained in each sub-picture is contained in the same consistent cropping window in the picture, conditions for designing the plurality of sub-pictures include that a right end of each of the plurality of sub-pictures is located rightward of a left end of the consistent clipping window.

The application discloses a kind of based on semantic perception's image cropping method and device, which effectively makes up the limitation of traditional method only relying on visual features, by introducing concept dictionary to enrich semantic information;Meanwhile, the present application further excavates potential semantic relationship through semantic similarity, which is helpful to fully exploit and utilize semantic information, so that a more rich and unified concept space is established between image and semantic label, further improving the accuracy of cropping result and the robustness of model;In addition, the present application breaks through the limitation of fixed angle of traditional method, by comprehensively considering the angle information of each instance, dynamically adjusting the direction of cropping frame, and by comprehensively considering the geometric properties, direction information, density distribution and other features of instance, automatically selecting appropriate cropping mode and composition ratio, so that the cropping result can more flexibly and accurately adapt to the layout and structure of image content.

The application discloses a kind of unmanned aerial vehicle adaptive control system for image cropping, comprising: unmanned aerial vehicle camera attitude control module, unmanned aerial vehicle camera attitude control module controls unmanned aerial vehicle camera and the midpoint of measured photovoltaic panel is in relative vertical state;Unmanned aerial vehicle camera attitude control module also controls the flight height and coordinate of unmanned aerial vehicle according to standard vertical relative height;Standard vertical relative height is theoretical value, and is determined according to the actual parameter of measured photovoltaic panel pixel size, known unmanned aerial vehicle camera parameter and measured photovoltaic panel in system setting;Image cropping module, image cropping module carries out image cropping according to the standard vertical relative height of unmanned aerial vehicle camera shooting image, the real-time vertical height of shooting image and the measured photovoltaic panel pixel size in system setting;The application has the beneficial effect that the image difference of different photovoltaic panels obtained is small, and cropping is more convenient.

The present application relates to the technical field of target detection and template matching, in particular to a workpiece recognition and counting method based on target detection and template matching in a complex scene, comprising the following steps: obtaining a standard workpiece original data set by using a high-definition industrial camera; obtaining an enhanced data set by pre-processing, data augmentation, cropping and classification labeling of the workpiece image set; constructing a target detection network model based on an improved YOLOv5 and performing data set training; deploying the trained target detection model to read the industrial camera in real time to detect the workpieces on the production line, crop and save the recognized workpiece images after rotating the angle. The present application has the advantages that: firstly, the workpieces are recognized from the complex scene image by the deep learning target detection method, and the position information of the workpieces is obtained, which has better robustness for the sensitivity of environmental information. It can handle multi-target and multi-angle recognition tasks, and can excellently complete workpiece recognition and counting in a relatively complex production scene.

A first object detection section (30) detects an object from a captured image and outputs object position information in a non-fast processing mode. A switching section (20) switches to a fast processing mode in a case where the object position information is output from the first object detection section (30). An image cropping section (40) extracts a cropped image from the captured image based on the object position information output from the first object detection section (30) in the fast processing mode. A second object detection section (60) detects an object from the cropped image and outputs object position information. An extension determination section (70) determines whether the object detected by the second object detection section (60) extends from the cropped image. The switching section (20) switches to the non-fast processing mode in a case where it is determined that the object detected by the second object detection section (60) extends from the cropped image.

The application provides an image dynamic cropping and frame buffer method and device, and relates to the technical field of image processing. The method comprises the following steps: in response to an image display task, acquiring a to-be-processed image frame sequence, cropping requirement information and display adaptation information; determining a buffer write control rule and a cropping read control rule according to the cropping requirement information and the display adaptation information; alternately writing the image frame into a first buffer area and a second buffer area based on the buffer write control rule; when the image frame is written in one buffer area, reading buffer frame data corresponding to a target cropping area from another buffer area in which the image frame has been written to obtain cropping frame data; and performing output adaptation processing on the cropping frame data according to the display adaptation information to obtain display image data. Through the method provided by the application, the cropping area adjustment flexibility can be improved, the image buffer and cropping processing efficiency can be improved, and the adaptability between the cropped image and the display output can be improved.

This application provides a mobile robot task scheduling method and system based on a heterogeneous computing architecture, relating to the field of intelligent scheduling technology. The method includes: constructing a workspace geometric model based on the geometric structure of a planting row; obtaining the motion state parameters of the mobile robot using a first computing unit; determining the look-ahead compensation amount by combining the statistical value of the inference time of the target recognition task by a second computing unit; determining a spatiotemporal pre-aiming slice in the visual field region based on the look-ahead compensation amount and the workspace geometric model, and cropping image data located within the slice from the original image data; the second computing unit performing target recognition on the image data, generating a work instruction corresponding to the recognized target, and controlling the lateral actuator to perform the work according to the work instruction. This can alleviate the spatiotemporal asynchrony problem between perception results and execution position in a moving work scenario, reduce missed and misaligned tasks, and reduce the communication and computational load of the heterogeneous computing system.

The invention discloses a boiler expansion monitoring method based on key point auxiliary detection, and belongs to the field of boiler expansion monitoring, and the method comprises the following steps: installing a predefined graphic target at a boiler expansion monitoring point, the graphic target is composed of a plurality of triangles and rectangles in a matrix arrangement mode, each pattern and the angular points thereof are numbered in the clockwise direction; shooting the graphic target through an industrial camera to obtain a target image; positioning a target area in the target image by using a target detection model and cutting the target area; key point detection is carried out on the cut target area, and angular points of a target pattern are identified and numbered; determining a centroid coordinate of each pattern through a corner point index and a polygon approximation algorithm; and based on the center-of-mass coordinates, calculating the three-dimensional pose of the target by using a PnP algorithm, and outputting boiler expansion displacement data. The positioning error of the method in an industrial scene is reduced by more than 60% compared with that of a traditional method.

This application provides a method, system, device, and storage medium for identifying smoking behavior in industrial parks. The method includes: acquiring video stream data from park surveillance; decoding the video stream data to obtain image frames; performing human target detection on the image frames to determine the location of each human body within the image frames; cropping sub-images containing the human bodies from the image frames based on their locations; performing smoking behavior recognition on the sub-images to determine whether the human body in the sub-image is smoking; and outputting the smoking behavior recognition results for each human body in the image frames. This method, through a two-stage processing architecture of first human detection and then region cropping, transforms the problem of identifying small-target smoking behavior in a large-scale park surveillance scenario into a refined behavior discrimination problem at a normalized scale. This effectively alleviates the contradiction between sparse features of small targets and background interference, improving the balance between recognition accuracy and computational efficiency.

This invention relates to the field of document image processing, proposing a method and system for batch cropping of stitched images based on segmentation line detection and quality verification. The method acquires an image formed by stitching multiple pages of content row by row and column, initializes the ultra-high resolution image, performs grayscale conversion and noise reduction, performs edge detection, and enhances linear structures using directional morphological operations; based on the enhanced results, candidate line segments are obtained through line detection, and segmentation line coordinates are filtered by directional constraints and boundary exclusion, and similar segmentation lines are merged, deduplicated, and sorted; cropping regions are constructed based on ordered segmentation lines to generate candidate sub-images, and invalid slices are filtered through minimum size / area thresholds and geometric consistency verification, outputting a verification list when necessary; the numbers of sub-images that pass verification are saved, and anomaly information is recorded. Through the above scheme, the cropping consistency and processing stability of automatic splitting of stitched document images are improved, erroneous cutting is reduced, and the traceability of results is enhanced.

1. The name of the design product: editing pop-up window of dynamic expression editing graphical user interface for electronic device. 2. The use of the design product: an electronic device. 3. The design points of the design product: the part in the graphical user interface, the part outside the dashed line is the claimed part. 4. The picture or photo that best shows the design points: design 1 front view. 5. Design 1 is designated as the basic design. 6. The use of the graphical user interface: the overall design of the interface is used for expression function setting, showing dynamic expression editing pop-up window, cropping expression, selecting dynamic expression key frame, etc., and the partial appearance design of the interface is used for showing dynamic expression editing pop-up window, synchronous cropping and selecting key frame expression online editing function, etc. 7. The human-computer interaction mode of the graphical user interface: in design 1 front view, the user can drag the cropping frame in the area below "crop expression" to crop the required expression picture; can click to select any dynamic key frame as a thumbnail, and can click "OK" control to complete editing. The human-computer interaction mode of design 2 to design 4 is the same as that of design 1. In design 5 front view, when the user drags the cropping frame in the area below "crop expression", the selected area is recropped, and the interface changes from design 5 front view to design 5 interface change state view. The area below "crop expression" in each design view is the content picture, such as dynamic expression, dynamic video, etc.; below the area of dynamic content picture, the key frame thumbnail content of dynamic content picture is displayed.