Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

404 results about "Good image" patented technology

Optical lens assembly and head-mounted electronic device

An optical lens assembly includes a phase retarder; a partial-reflective-partial-transmissive element; and in order from a visual side to an image source side: an absorptive polarizer, a reflective polarizer, a first lens with refractive power, a second lens with refractive power, and a third lens with positive refractive power; wherein the phase retarder and the partial-reflective-partial-transmissive element are sequentially arranged from the visual side to the image source side, and are disposed between the reflective polarizer and the third lens. The optical lens assembly may become lightweight and have good image quality when satisfying a specific condition.
Owner:NEWMAX TECH CO LTD

High-performance airborne head-mounted low-light night vision optical system

The invention relates to the technical field of optical design, in particular to a high-performance airborne head-mounted low-light night vision optical system which sequentially comprises a first positive lens, a first gluing assembly, a fourth positive lens, a fifth positive lens, a second gluing assembly and an eighth negative lens from an object side to an imaging surface along an optical axis. The image side surface is a concave surface; the first gluing assembly has a concave object-side surface and a concave image-side surface. The fourth positive lens has positive focal power, the object side surface of the fourth positive lens is a convex surface, and the image side surface is a convex surface; the fifth positive lens has positive focal power, the object side surface of the fifth positive lens is a convex surface, and the image side surface is a convex surface; the second gluing assembly has a concave object-side surface and a convex image-side surface; the eighth negative lens has negative focal power, the object side surface of the eighth negative lens is a concave surface, and the image side surface is a convex surface. The optical system has the advantages of large relative aperture, low cost, good image quality, small size, light weight and good thermal stability, and can work normally at low temperature and high temperature.
Owner:HUNAN AEROSPACE JIECHENG ELECTRONIC EQUIP CO LTD

Wide-angle video monitoring system, method and equipment based on multiple cameras and medium

The invention discloses a wide-angle video monitoring system, method and equipment based on multiple cameras and a medium, and relates to the technical field of video monitoring, the wide-angle video monitoring system comprises a multi-camera acquisition module, a variable-focus high-definition camera multi-angle acquisition module, a high-quality image obtained through distortion correction and an image splicing processing module, an improved algorithm is used for accurately matching feature points and fusing the images, and the high-quality image is obtained. The intelligent analysis module detects and tracks a target and analyzes behaviors in real time by means of an improved model, the storage and transmission module performs cluster storage, advanced coding cost reduction and high-speed link transmission, and the control module performs intelligent regulation and control according to multiple information and can also perform remote management. The system has the advantages of multi-angle acquisition without dead angles, good image splicing, accurate intelligent analysis, efficient storage and transmission, adaptability to severe environments, flexible power supply, guarantee of data safety, improvement of monitoring quality and efficiency, meeting of complex scene requirements, and enhancement of practicability, reliability and safety.
Owner:CHANGZHOU ZUOAN ELECTRIC CO LTD

Warehouse goods intelligent identification and positioning management system based on image segmentation algorithm

A warehouse goods intelligent identification and positioning management system based on an image segmentation algorithm comprises an image acquisition and preprocessing module, an image segmentation algorithm module, a goods positioning and feedback module, an inventory management module, a data communication and integration module and a user interaction and management module. The image segmentation algorithm module is used for segmenting and extracting image information. The cargo identification and positioning module is used for identifying and positioning cargoes. The position feedback and alarm module is used for feeding back cargo positions and giving an alarm. The inventory management module is used for managing and maintaining cargo information. The data communication and integration module is used for transmitting information in real time. According to the warehouse goods intelligent identification and positioning management system based on the image segmentation algorithm, the warehouse image segmentation algorithm based on deep learning is put forward to classify goods, and the goods identification and positioning algorithm based on improved label template matching is put forward to intelligently position the goods.
Owner:ZHONGSHAN DIANNIU TECH CO LTD

In-vehicle conferencing with user tracking and speaker prioritization

Systems and methods for implementing smart in-vehicle conferencing with user tracking and speaker prioritization. A controller system of a vehicle may determine, based at least in part on one or more cameras, a plurality of candidate best image sources, determine, from the plurality of candidate best image sources, a best image source for a speaker in the vehicle, determine, based at least in part on the best image source, a video feed, determine, based at least in part on one or more microphones, a plurality of candidate best audio sources, determine, from the plurality of candidate best audio sources, a best audio source for the speaker, determine, based at least in part on the best audio source, an audio feed, and provide, to a conference call line, the video feed and the audio feed
Owner:FORD GLOBAL TECH LLC

Start-stop two-way linkage machine vision detection method for longitudinal tearing of conveying belt

The invention discloses a start-stop bidirectional linkage conveyor belt longitudinal tearing machine vision detection method, and relates to the technical field of coal-electricity conveyor belt safety operation and maintenance, and the method comprises the steps: obtaining a bottom longitudinal tearing condition feature image of a conveyor belt in a running state; frame-by-frame image extraction, image filtering, deformation detection, image relevance decision and mask processing are carried out on the bottom longitudinal tearing condition feature image, an alarm signal and a bottom longitudinal tearing condition feature image after abnormal processing are generated, and the alarm signal and the bottom longitudinal tearing condition feature image after abnormal processing are acquired; system linkage is realized; according to the method, after system deployment and debugging are completed, an alarm threshold value is set in advance according to the running speed of the belt, all-weather continuous image recognition is conducted, automatic grading alarm is conducted when abnormal features of the surface of the belt are detected, screenshots are recorded in a database, and long-time training is not needed; the method is good in image effect, high in tearing recognition rate, high in algorithm accuracy and capable of being stably used for a long time.
Owner:SHANDONG RONGXIN IOT TECHNOLOGY CO LTD

Foggy day pedestrian image enhancement method based on CNN multi-scale feature fusion

The invention discloses a foggy-day pedestrian image enhancement method based on CNN multi-scale feature fusion, and the method comprises the steps: constructing a foggy-day pedestrian image enhancement model which is of a U-shaped network structure and comprises an encoder module, a feature fusion module and a decoder module; in the encoder module and the decoder module, lightweight FastCNN modules are adopted, so that the parameter quantity and the calculation quantity of the model can be effectively reduced; the feature fusion module comprises a CSP-ADD module, and the CSP-ADD module enhances the feature extraction capability by stacking a plurality of residual blocks, effectively transmits gradients in a deep neural network, keeps detail information of input features, and improves the focusing capability of the network on important information, thereby further enhancing the effects of tasks such as defogging and denoising. The model constructed by the method has the advantage of light weight, and meanwhile, a relatively good image restoration effect can be ensured.
Owner:HUAIYIN INSTITUTE OF TECHNOLOGY

Photon flow detector with high image acquisition performance and CT equipment

The invention discloses a photon flow detector with high image acquisition performance and CT (Computed Tomography) equipment. The photon flow detector comprises a scintillator, a photon flow sensor, a main control board and a substrate. Wherein the scintillator is in coupling connection with the photon flow sensor, and the output end of the photon flow sensor is connected with the main control board. The scintillator is used for converting incident X photons into visible light photons; the photon flow sensor is used for converting visible light into an electric signal, generating an integral signal and a counting signal through amplification shaping, integration and counting processing, and outputting image data after data conversion and processing; the main control board is used for providing a clock signal, a working power supply and data analysis and transmission. When the photon flow detector carries out detection work, the photon flow sensor generates an integral signal and a counting signal at the same time, image data is formed after data conversion and processing, the image data is uploaded to the image acquisition equipment, and good image acquisition performance is achieved in the whole dose range.
Owner:NANOVISION TECHNOLOGY (BEIJING) CO LTD

Optical lens assembly and electronic device

An optical lens assembly includes, in order from a visual side to an image source side: an optical group including a partial-reflective-partial-transmissive element; and in order from the visual side to the image source side: an absorptive polarizer, a reflective polarizer, a phase retarder, a first lens, a second lens, and a third lens; wherein the partial-reflective-partial-transmissive element is disposed between the first lens and the third lens; at least one of the first lens and the second lens has a positive refractive power; when the focus of the optical lens assembly changes from a near point to a far point, the optical group moves from the image source side to the visual side. The optical lens assembly may become lightweight and have good image quality when satisfying a specific condition.
Owner:NEWMAX TECH CO LTD

Image deblurring model based on double-domain feature fusion and method thereof

The invention relates to the technical field of image processing, and discloses an image deblurring model and method based on double-domain feature fusion, and the method comprises the following steps: collecting data, and building a blurred image data set; building a DAF-UNet neural network based on spatial domain and frequency domain feature fusion; the DAF-UNet neural network is composed of n UNet sub-networks with different scales, and n is a positive integer; based on the blurred image data set, training the DAF-UNet neural network to obtain a trained image deblurring model; collecting a to-be-processed image; and processing a to-be-processed image by using the trained image deblurring model to obtain a clear image result. According to the method, the problems of low information mobility and key feature loss in the prior art are solved, and the method has the characteristics of good image recovery effect and model miniaturization.
Owner:GUANGDONG POLYTECHNIC NORMAL UNIV

Automatic steel belt deviation rectifying method based on machine vision

The invention relates to the field of steel belt monitoring, in particular to a steel belt automatic deviation correction method based on machine vision, and the method comprises the steps: obtaining an image when a steel belt works, calculating the self-adaptive standard deviation of each pixel point in the image in a bilateral filtering algorithm, and carrying out the denoising of the image through the bilateral filtering algorithm, so as to obtain an optimal image; performing edge detection on the optimal image to obtain the edge of the steel belt, calculating the distance between the edge of the steel belt and the standard position, and adjusting a guide roller according to the obtained distance to enable the steel belt to be located in the standard position range; the position identification accuracy of the steel belt can be improved, and deviation correction of the steel belt is facilitated.
Owner:WUXI COFCO ENG & TECH CO LTD

Blind image restoration method and system, electronic equipment and storage medium

The embodiment of the invention provides a blind image restoration method and system, electronic equipment and a storage medium. The method comprises the following steps: acquiring a to-be-restored image; inputting the to-be-restored image into a pre-established blind image restoration model, and outputting to obtain a restored image; wherein the blind image restoration model is pre-established based on a vector quantization variational auto-encoder, a multi-layer perceptron and a continuous diffusion generation model. According to the embodiment of the invention, by utilizing the superiority of the continuous diffusion generation model in learning data manifold distribution, the vector quantization variational automatic encoder is used as a codec between the image space and the submerged space, and the generation process in the submerged space of image encoding is constructed, so that a low-quality image under complex degradation is repaired, and the image quality is improved. And a good image restoration effect is achieved. A continuous diffusion generation mode is introduced, and a preprocessed guide embedding mode is constructed, so that guide conditions in a high-quality image prediction process are more reasonable and interpretable.
Owner:UNIV OF CHINESE ACAD OF SCI

Robust reversible watermarking method based on BFMs and adaptive normalization

The invention discloses a robust reversible watermarking method based on BFMs and adaptive normalization. The method comprises the steps of watermark embedding and watermark extraction. The watermark embedding specifically comprises the steps of generating watermark sequence information, performing integer wavelet transform on a carrier image, extracting a low-frequency region, and calculating BFMs of the low-frequency region; a proper watermark pair is selected through a private key, self-adaptive normalization is carried out according to the watermark pair, and the watermark is embedded in the normalized amplitude; reconstructing an image based on the amplitude of the embedded watermark; constructing a final reversible watermark image based on the reconstructed image; watermark extraction: judging whether the image is attacked or not; for an unattacked image, carrying out watermark extraction based on the auxiliary information; for an attacked image, watermark extraction is carried out by calculating the BFMs of the attacked image. The method has noise resistance and also has relatively good image reconstruction capability.
Owner:NANJING UNIV OF INFORMATION SCI & TECH

Automatically generating best digital images of a person in a physical environment

The disclosed technology provides for generating best images of a person in a retail environment. A method may include receiving, by an edge computing device from a camera, a continuous stream of image data of the retail environment, detecting, using object detection techniques, a person in the image data, the image data including a group of images that are part of a time series, generating bounding boxes for each of the group of images around the person based on detecting the person as they move in the images, identifying, based on applying a features model to each bounding box, at least one feature of the group of images depicting the person, selecting a subset of the bounding boxes having at least one feature that satisfies best images criteria, the subset having best images of the person, and returning the best images of the person.
Owner:TARGET BRANDS INC

Aerospace area-array camera in-orbit multi-stage real-time automatic dimming device and method

The invention relates to the technical field of image processing, in particular to an in-orbit multi-stage real-time automatic dimming device and method for a spaceflight area-array camera. According to the device, an imaging module obtains an original scene image according to set integral time and transmits image data to an image coarse adjustment module; the image coarse adjustment module counts gray level distribution of an original image in real time, calculates first updating integral time, packages image data without overexposure or underexposure and sends the packaged image data to the image fine adjustment module; after the image fine adjustment module receives the image, scene recognition and gray level evaluation are carried out through a deep learning model, exposure parameters are dynamically adjusted according to the scene type, and image exposure control is achieved; and the image coarse adjustment module and the image fine adjustment module form a two-stage closed loop, so that the imaging module obtains integral time matched with a scene in real time. The method has the advantages that the integral time of the detector is automatically adjusted according to the detail content of the target scene, and the target image with the best image impression and detail is captured.
Owner:CHANGCHUN INST OF OPTICS FINE MECHANICS & PHYSICS CHINESE ACAD OF SCI

Endoscope objective lens, endoscope, and endoscope system

The invention relates to an endoscope objective lens. The endoscope objective lens sequentially comprises a first lens and a first balsaming lens from an object plane to an image plane. The first bonding lens moves on the optical axis of the endoscope objective lens, so that the difference between the focal length of the first bonding lens in the endoscope objective lens and the value of the movement of the first bonding lens on the optical axis of the endoscope objective lens is between 15.00 and 35.00; according to the invention, the focal length of the first balsaming lens and the distance of far and near focus switching of the first balsaming lens can be properly proportioned, the aberration correction and the tolerance sensitivity of the moving group are facilitated, the total length is short, the length of the hard part of the whole endoscope is reduced, the observation depth of the endoscope can be improved, and the service life of the endoscope is prolonged. A good image observed in a close range is ensured, and the accuracy of observation and information acquisition of the endoscope is effectively improved.
Owner:CHANGZHOU UNITED IMAGING HEALTHCARE SURGICAL TECH CO LTD

Stereoscopic warehouse stacked goods checking method based on improved YOLO11-Seg

The invention provides a stereoscopic warehouse stacked goods checking method based on improved YOLO11-Seg. The stereoscopic warehouse stacked goods checking method comprises the steps of collecting goods images and performing labeling processing on the goods images; the method comprises the following steps: constructing an improved YOLO11-Seg model; carrying out segmentation identification on the cargo image by utilizing the trained improved YOLO11-Seg model; judging whether the condition is legal or not according to the segmentation mask and the mask category information; according to the geometric relationship of the convex hull points, judging whether the cargo top layer is in a full load condition, and if the cargo top layer is in the full load condition, directly calculating the cargo quantity; otherwise, calculating the number of the sub-top-layer boxes according to an estimation algorithm, and then calculating the total number of the cargo boxes. According to the method, the identification of stacked goods is realized by utilizing an improved YOLO11-Seg model, a goods segmentation mask and a mask type are generated, and the total goods quantity is calculated by identifying and distinguishing different layers of goods boxes; the method has the advantages of high generalization, low deployment cost, high checking efficiency and the like.
Owner:GUANGDONG UNIV OF TECH

Weak supervision directivity segmentation method based on semantic and detail cooperation

The invention belongs to the technical field of artificial intelligence, and discloses a weak supervision directivity segmentation method based on semantic and detail collaboration. The method comprises the following steps of: firstly, respectively extracting features of an input image and a text by utilizing two encoders based on Transform, and then inputting the two features into a semantic perception module and a detail cognition module; the two modules respectively pay attention to high-level semantics and low-level details and are used for a collaborative learning strategy, and the strategy integrates cross-modal internal loss, matching invariance loss and regional contrast loss to achieve sufficient visual language alignment. The semantic perception module and the detail cognition module respectively generate an activation graph, and then the activation graphs are integrated by a cooperation module to output accurate segmentation masks. According to the method, only the image text is used for training the supervision model, the method does not depend on dense pixel-level labels, and good image segmentation performance indicated by natural languages is achieved while the manual labeling cost is remarkably reduced.
Owner:DALIAN UNIV OF TECH

A method and system for fusing laser point cloud and visible light image

The first aspect of this application proposes a method for fusing laser point cloud and visible light image, including: S1, performing spatial dimensionality reduction on the point cloud data to obtain two-dimensional point cloud data; S2, extracting feature points of the two-dimensional point cloud data and feature points of the visible light image data through the Scale-Invariant Feature Transform (SIFT) algorithm; S3, determining the projection plane, projecting the point cloud data through the projection model to generate a depth-echo intensity image, and constructing a 3D-2D coordinate mapping relationship between the point cloud data and the depth-echo intensity image; S4, mapping the feature points of the visible light image data to the point cloud data according to the 3D-2D coordinate mapping relationship. This solution can achieve a good image registration and fusion effect, can be applied to various platforms, and has a wide range of applicable scenarios. The second aspect of this application proposes a system for fusing laser point cloud and visible light image.
Owner:FUJIAN STRAIT ZHIHUI TECH CO LTD

A continuously variable magnification prism telescope optical system

The present invention discloses a continuously variable magnification prism telescope optical system, belonging to the technical field of optical telescope systems. The system comprises a first objective lens doublet, a third objective lens, a second objective lens doublet, a third objective lens doublet, a fourth objective lens doublet, a tenth objective lens, a first prism, a graticule, an eyepiece doublet, a fourth eyepiece lens, and a fifth eyepiece lens, arranged in sequence along a first optical path. Most of the objective lenses and eyepieces in the continuously variable magnification prism telescope optical system are doublets, which can effectively shorten the overall length of the optical system and reduce the number of lenses in the optical system, resulting in a smaller volume and weight of the entire optical system after assembly, while also achieving good image quality. The system achieves continuous changes in focal length and / or imaging magnification by moving the second objective lens doublet and / or the third objective lens doublet along the optical axis. The use of a variable magnification objective lens structure makes the eyepiece structure simple and highly reliable.
Owner:HUBEI HUAZHONG PHOTOELECTRIC SCI & TECH CO LTD +1

Coding method, signal processing method and related equipment

The signal processing method provided by the embodiment of the invention comprises the steps that a first basic image, a first gain graph and metadata are acquired, and the first basic image corresponds to a first dynamic range; processing the first basic image according to the metadata to obtain a second basic image; according to the second basic image and the first gain image, a target image is obtained, the target image corresponds to a second dynamic range, and the second dynamic range is different from the first dynamic range. The scheme provided by the invention can be compatible with coding and processing of images in various formats, and relatively good image quality can be obtained on systems with different support capabilities.
Owner:HUAWEI TECH CO LTD

Millimeter wave up-down conversion module

ActiveCN119093963AGood image rejection ratioLow vibration leakageTransmissionHigh level techniquesLocal oscillator signalHybrid coupler
The invention discloses a millimeter wave up-down conversion module, which comprises an intermediate frequency transmit-receive change-over switch and a radio frequency transmit-receive change-over switch, and two links, namely an up-conversion link and a down-conversion link, are arranged between the intermediate frequency transmit-receive change-over switch and the radio frequency transmit-receive change-over switch. The up-conversion link sequentially comprises an intermediate frequency amplifier, a hybrid coupler and a quadrature up-converter. The down-conversion link includes a down-converter. In addition, a local oscillator circuit provides local oscillator signals for the quadrature up-converter and the quadrature down-converter. And a radio frequency balun transformer II is connected with the radio frequency transmit-receive change-over switch. The millimeter wave up-down conversion module has two working modes, namely a transmitting mode and a receiving mode. The signal passes through the up-conversion link in the transmitting mode, and the signal passes through the down-conversion link in the receiving mode. The up-down conversion module has the advantages of good image rejection ratio, low local oscillation leakage, broadband matching, small area and low design complexity.
Owner:ASR MICROELECTRONICS CO LTD

In-vehicle conferencing with user tracking and speaker prioritization

Systems and methods for implementing smart in-vehicle conferencing with user tracking and speaker prioritization. A controller system of a vehicle may determine, based at least in part on one or more cameras, a plurality of candidate best image sources, determine, from the plurality of candidate best image sources, a best image source for a speaker in the vehicle, determine, based at least in part on the best image source, a video feed, determine, based at least in part on one or more microphones, a plurality of candidate best audio sources, determine, from the plurality of candidate best audio sources, a best audio source for the speaker, determine, based at least in part on the best audio source, an audio feed, and provide, to a conference call line, the video feed and the audio feed.
Owner:FORD GLOBAL TECH LLC

Camera shooting control method and device using optical fiber signal, equipment and medium

The invention discloses a camera shooting control method and device using an optical fiber signal, equipment and a medium. After the system is powered on, the upper computer sends control commands such as communication and triggering, the control commands are converted into optical signals through the electro-optical conversion module, the optical signals are sent to controlled equipment (including a camera unit, a light source unit and an aperture unit) through optical fibers (plastic optical fibers and silica optical fibers), the optical signals are converted into electric signals through the photoelectric conversion module, and the system works based on the control commands. Operation such as camera exposure, light source flashing and aperture opening and closing is achieved, and ideal control states of good image sensitization, light source synchronization and moderate aperture are achieved. Compared with a traditional control mode that electric signals are transmitted through cable media, optical signals are transmitted through the optical fiber media, the electromagnetic interference influence can be reduced, the transmission distance is longer, the cable weight is lighter, the error control probability is reduced, and the overall system robustness is effectively improved.
Owner:MICROVIEW SCI & TECH

Cargo quantity checking method and device based on cross validation, equipment and medium

The invention discloses a cargo quantity checking method and device based on cross validation, equipment and a medium, and the method comprises the steps: numbering cargoes to be checked in a stack, and collecting the images of the cargoes in the stack; carrying out stack type classification and identification on the goods image before taking the goods, and establishing a position state matrix; carrying out positioning detection on the goods image before the goods are taken, and generating shielding information and a confidence score; acquiring depth information of the image of the cargo in carrying, updating the position state matrix, and generating shielding information and a confidence score; carrying out positioning detection on the cargo image after cargo placement, updating the position state matrix, and generating shielding information and a confidence score; and performing cross validation in combination with the shielding information, generating target shielding information, performing cross validation in combination with the confidence score, generating a target confidence score, and generating an in-stack cargo quantity checking result according to the target confidence score and the target shielding information. According to the invention, the checking precision and the checking reliability of the number of goods in the stack can be improved.
Owner:NEW TREND INT LOGIS TECH

Video image fusion processing method and device for capsule endoscope

The invention discloses a video image fusion processing method and device for a capsule endoscope, and the method comprises the steps: loading a plurality of segments of videos at different angles, which are shot by the capsule endoscope, into a processing module, carrying out the image enhancement preprocessing of the nth frame of each segment of video, carrying out the rough image registration of the preprocessed image through the pose data of a camera, and carrying out the image fusion processing; and carrying out image fine registration by using an SURF algorithm to determine an overlapping region, searching an optimal suture line in the overlapping region by using an optimal suture line image fusion algorithm, carrying out image stitching, and carrying out video stitching on the processed stitching image set. According to the invention, image coarse registration is carried out according to the pose data of the capsule endoscope, an SURF image registration algorithm is improved, and optimization and improvement are carried out in a feature point matching stage, so that the accuracy and timeliness of the algorithm are improved.
Owner:CHINA JILIANG UNIV

Palm acupoint selection moving system based on valley point positioning

The invention discloses a valley point positioning-based palm acupoint selection mobile system, which is characterized by comprising a mobile terminal and a palm acupoint positioning module, the mobile terminal is used for providing health preserving suggestions and feeding back acupoint selection results by identifying the hand photos online; and the palm acupuncture point positioning module is used for accurately positioning palm acupuncture points and providing a basis for providing health preserving suggestions in combination with a hand photo recognition result. On the premise that good image processing is achieved, palm acupoint selection and health maintenance are more convenient through the mobile terminal.
Owner:SHAANXI UNIV OF SCI & TECH

Automatic repairing method for ultrasonic image

The invention relates to the technical field of ultrasonic image processing, in particular to an automatic repairing method of an ultrasonic image, which comprises the following steps: identifying a to-be-repaired image to obtain a to-be-repaired area, and intercepting an interval echo sequence from a complete ultrasonic echo sequence for the to-be-repaired area; classifying the interval echo sequence and the to-be-repaired area to determine defect factors; the defect factors comprise at least one of scanning offset, vibration and medium change. And calling a corresponding mask plate according to the defect factors, and repairing the to-be-repaired area according to the mask plate. In order to solve the problem that in the prior art, when ultrasonic inspection equipment synthesizes an ultrasonic image of a large area, local defects exist, and artificial intelligence model training is affected, a process of extracting a to-be-repaired area of the local defects and analyzing defect factors of the to-be-repaired area according to an echo sequence is introduced, the defect factors causing the imaging defects are determined, and the defect repairing efficiency is improved. And the corresponding mask plate is called for repairing, so that a relatively good image repairing process is realized.
Owner:SHANGHAI SOUNDWISE TECHNOLOGY CO LTD

Warehouse goods image recognition method based on deep learning

The invention discloses a deep learning-based warehouse cargo image recognition method, and particularly relates to the technical field of warehouse cargo recognition, and the method comprises the steps: simulating the day and night illumination difference through an image enhancement module, and constructing a training data set covering a real scene; carrying out feature extraction by using an improved ResNet101 model, and reconstructing a 256-dimensional full-connection layer to output a high-discrimination cargo exclusive feature vector; a double-branch classification structure is innovatively adopted, the first branch judges the goods category, the second branch judges the packaging state, and according to a confidence coefficient threshold value, intelligent decision making of high-confidence-coefficient result automatic output and low-confidence-coefficient triggering rechecking is achieved; the system integrates an error analysis module, a performance evaluation module and a model dynamic optimization module, and supports scene-based directional incremental updating. The method overcomes the defect that the prior art is poor in adaptability to complex scenes, illumination changes and package anomalies, remarkably improves the recognition precision and the operation efficiency, and effectively reduces the inconsistency of inventory accounts.
Owner:JIANGSU TIANZHIJIE INFORMATION TECH SERVICE CO LTD

Extreme image coding and decoding method and device based on diffusion model

The invention discloses an extreme image coding and decoding method based on a diffusion model. The method comprises the following steps: S100, a compression and pre-decoding module at a coding end extracts, compresses and transmits image information by using a neural coding network; s200, a compression and pre-decoding module of the decoding end preliminarily decodes the received image information into a content variable aligned with the potential diffusion space by using a neural decoding network; and S300, a denoising module uses the content variable as a control condition and a text graph diffusion model as prior information, and reconstructs an image from random noise through step-by-step denoising. The method can utilize the strong generation capability of the pre-trained text image diffusion model to realize perception-friendly reconstruction consistent with the original image at an extremely low bit rate, has the characteristics of ultrahigh compression ratio and good image reconstruction effect, can carry out image transmission under an extremely low bandwidth condition, and has a wide application prospect. The method can be widely applied to the fields of smart phone satellite communication, short-wave communication, unmanned equipment remote operation and the like.
Owner:XI AN JIAOTONG UNIV