8results about How to "Excellent extractability" patented technology

The invention discloses a generative model redrawing attack-oriented image robust watermark traceability method and system and a computer readable storage medium, and belongs to the field of digital information security and artificial intelligence content governance. The method comprises the following steps: in a watermark embedding stage, performing error correction coding and digital signature processing on traceability information containing identity information and a timestamp to generate a load to be embedded, and embedding the load to be embedded and a synchronization template for geometric synchronization into a host image in a function separation manner by using a deep neural network; in the training stage, combined optimization is carried out on the watermark embedding and extracting process by introducing generative model redrawing and microsimulation of image distortion attack, so that the robustness of the watermark under a complex attack condition is improved; in the extraction and verification stage, under the condition that an original image is not needed, geometric synchronous correction and blind extraction of a watermark load are carried out on an image to be analyzed, and authenticity confirmation of traceability information is completed through error correction decoding and digital signature verification. By adopting the technical scheme of the invention, the problems of insufficient traceability information robustness, difficult source confirmation and incomplete evidence chain in a generative model redrawing scene in the prior art are solved, and verifiable traceability and credible evidence generation of the image source information are realized.

The invention belongs to the technical field of secret communication, and discloses an unstructured data processing method based on additional information embedding and homomorphic encryption. The method comprises the following steps: generating a feature vector by preprocessing unstructured data, generating an embedded signal by combining an additional information coding and modulation technology with the feature vector, and dynamically adjusting an embedding position through an optimization algorithm to minimize the influence on data quality so as to complete the embedding of additional information; encrypting the data embedded with the additional information by using an encryption algorithm supporting homomorphic operation; in the data use process, the embedded additional information is recovered through decryption, feature vector extraction, inverse modulation and decoding. Compared with the prior art, the method has the advantages that the additional information can be effectively embedded in the unstructured data, meanwhile, the security and privacy of the data are guaranteed by means of the homomorphic encryption technology, and the method is suitable for various data forms and application scenes.

The application provides a palladium ion extractant material and a preparation method thereof, and the pyridine-oxadiazole ligand is used as the extractant, most Pd 2+ in the aqueous phase can be extracted into the organic phase in a short time, and has a very high extraction rate; selective extraction separation of Pd 2+ under acidic conditions can be realized, the applicable range of acidity is wide, and excellent extraction capacity is achieved in the range of 0.2-4M nitric acid concentration; the extractant has excellent selectivity to Pd 2+ in the aqueous phase containing various metal ions; the pyridine-oxadiazole extractant has a good application prospect in the separation and recovery of Pd 2+ in high-level radioactive waste; the m-nitro-trifluorotoluene is used as a diluent of the extraction system, has good solubility to the organic ligand and the extractant, and the 2-bromohexanoic acid can increase the lipophilicity of the extractant.

The invention provides a video stream data security enhancement method and device, electronic equipment and a storage medium, and relates to the technical field of data processing.When plaintext data is steganographically written into original video stream data, an edge detection mode and a stream encryption algorithm based on Gaussian filtering and Laplacian transformation are adopted, according to the method, the edge pixel points in the video stream data can be accurately detected based on an edge detection mode of Gaussian filtering and Laplacian transformation, and the positions of the detected edge pixel points do not change before and after the target encrypted data is embedded in the original video stream data, so that the visual effect of the video stream data is not influenced, and the accuracy of the video stream data is improved. According to the method, the embedded data is not easy to perceive, meanwhile, a receiver can accurately extract the embedded plaintext data from the target video stream data, and the stream encryption algorithm can enhance the security of the video stream data, so that the security and extractability of the hidden sensitive data in the video stream data are improved.

The utility model discloses an air inlet filtering mechanism for low-pressure casting, which comprises a connecting seat, a heating device arranged at the bottom of an inner cavity of the connecting seat, a liquid storage box arranged at the top of the heating device, a liquid inlet pipe arranged on the right side of the liquid storage box, a liquid delivery pipe arranged in the inner cavity of the liquid storage box, and a gas compression device arranged on the left side of the connecting seat. A supporting plate is arranged in an inner cavity of the connecting base, a lower die is arranged at the top of the supporting plate, and a hydraulic cylinder is arranged at the top of the connecting base. The low-pressure gas can be filtered through the filter screen when the low-pressure gas is blown out through the gas outlet pipe, the conditions of impurities and dust are prevented, meanwhile, an activated carbon material can adsorb small impurities of the gas, and therefore double filtration and purification can be conducted on the low-pressure gas, the quality of the low-pressure gas is improved, and the influence on the quality of a cast mold is prevented; and the output shaft of the electric motor drives the fan blades to rotate to blow and dissipate heat of the cast mold, so that the mold is quickly cooled, the time is shortened, and the working efficiency of a user is improved.

The invention discloses a foreign body removal device for otolaryngology, which relates to the technical field of otolaryngology, and comprises a pressing plate under a fixed block, a handle at one side of the fixed block, an arc plate on the fixed block, and an adjusting device for adjusting the height of the arc plate between the arc plate and the fixed block; an outer rod is arranged on the other side of the fixed block, an inner rod is arranged to slide inside the outer rod, one end of the inner rod slides through the side wall of a connecting pipe, a first slide groove is arranged at one end of the inner rod that penetrates the side wall of the outer rod, an illuminating lamp and a camera are respectively arranged on the inner rod on both sides of the first slide groove, two clamps are respectively arranged to slide in the first slide groove, and an electric telescopic rod is arranged between the clamp and the inner wall of the first slide groove; the invention has a simple and reliable structure, a good effect on removing foreign bodies, is easy to control, and is suitable for popularization.

The invention discloses a plane target adapting to a mobile scanning environment and a bull's-eye positioning method and system. Relates to the field of three-dimensional laser scanning and the technical field of engineering measurement. The method comprises the following steps: acquiring point cloud data of a complete front pattern of each planar target and preprocessing the point cloud data to obtain a target point cloud data set; processing the target point cloud data set to obtain a centralized point cloud, and performing singular value decomposition on the centralized point cloud; constructing a point cloud data plane according to the decomposition result of the singular value, calculating the distance from the point cloud data to the point cloud data plane, and determining an optimal plane; a cross wire of each plane target is decomposed, original data of straight lines are extracted and optimized, the optimized original data are fitted, and two straight lines are obtained; and projecting the two straight lines to an optimal plane to calculate an intersection point, thereby completing target center identification. The method can effectively remove the influence of noise and outliers, enables the data to be more concentrated and accurate, and remarkably improves the precision and reliability of subsequent calculation.

This invention relates to the field of sample pretreatment technology, and in particular to a rapid extraction method for pesticides in water based on a confined liquid membrane. The method includes the following steps: preparing a composite substrate membrane comprising carbon nanofibers / carbon fibers and carbon nanofibers / carbon spheres; injecting ethanol as a confining solvent through the composite substrate membrane until droplets form beneath the membrane, thus completing the construction of the confined liquid membrane; passing a water sample containing the target pesticide through the confined liquid membrane at a set flow rate; using a desorption solvent at a set flow rate through the confined liquid membrane to elute the extracted target pesticide; collecting the eluent and performing quantitative analysis. This invention, through its ingenious confined liquid membrane design, successfully achieves rapid simultaneous extraction of pesticides with a wide polarity range and multiple categories from water while ensuring high extraction efficiency and good reproducibility.