37 results about "Fingerprint retrieval" patented technology

The invention discloses a video fingerprint retrieval method based on coarse and fine granularity. According to the method, a video fingerprint is subjected to retrieval comparison in different-layer and different granularity retrieving mode. A local sensitive Hash algorithm is used for carrying out Hash processing on all video fingerprints in a video fingerprint database and the video fingerprint to be retrieved, so that the video fingerprint most similar to the video fingerprint to be retrieved is found, and the coarse granularity finding is completed, and the preparation is made for effectively reducing the time expenditure of the video retrieval; during the fine granularity finding, through the fast video fingerprint matching mechanism improved based on a biological sequence comparison technology BLAST, whether the video fingerprint to be retrieved exists in the video fingerprint database or not is fast found, the video segment matching finding can also be carried out, and in addition, the affiliated complete videos of the video segments to be retrieved and the concrete occurring time positions of the video segments to be retrieved in the complete videos can be determined according to the features of the video fingerprint, so that the accuracy and the real-time performance of the video finding are ensured.

The invention discloses an audio fingerprint retrieval method and a retrieval device. The audio fingerprint retrieval method includes the steps of carrying out framing process to audio signals of an audio clip, determining key frames in each subframe obtained through the framing processing by means of an onset detection function (ODF) method, obtaining audio fingerprints of the key frames, carrying out retrieval according to the obtained audio fingerprints, and obtaining audio file information corresponding to the audio clip. Correspondingly, the invention further discloses an audio fingerprint storage method and a storage device. The audio fingerprint storage method and the storage device can greatly reduce wasted memory due to audio retrieval, greatly reduce wasted time due to the audio retrieval, and improve retrieval efficiency on the premise of ensuring accuracy rate of audio fingerprint retrieval.

The invention relates to a method and system for updating an audio fingerprint search library. The method comprises the following steps of: storing audio fingerprints and retrieval indexes of all old songs in a basic search library; updating audio fingerprints and retrieval indexes of new songs in an incremental search library; judging whether the volume of the incremental search library is more than or equal to a preset threshold, if so, combining the incremental search library with the basic search library and generating a new basic search library, and updating the audio fingerprints and the retrieval indexes of the new songs in a new incremental search library; and, if not, updating the audio fingerprints and the retrieval indexes of the new songs in the incremental search library. According to the method and system disclosed by the invention, aiming at the condition that new songs are frequently added, when the incremental search library is very small and the cost for updating theincremental search library is very low, the incremental search library is only updated at each time, therefore, the time for updating the audio fingerprint search library is saved; and the audio fingerprint search library can be updated rapidly.

Methods are provided for both single modal and multimodal fault diagnosis. In a method, a fault fingerprint is constructed based on a fault event using an invariant model. A similarity matrix between the fault fingerprint and one or more historical representative fingerprints are derived using dynamic time warping and at least one convolution. A feature vector in a feature subspace for the fault fingerprint is generated. The feature vector includes at least one status of at least one system component during the fault event. A corrective action correlated to the fault fingerprint is determined. The corrective action is initiated on a hardware device to mitigate expected harm to at least one item selected from the group consisting of the hardware device, another hardware device related to the hardware device, and a person related to the hardware device.

The invention provides a fingerprint retrieval method based on the global direction information. The fingerprint retrieval method comprises the following steps: A, establishing a fingerprint global direction information feature representation model; B, extracting the global direction information of the fingerprint database and the input fingerprint image, using the global direction information feature to represent the model to construct the feature vector, and then performing searching. The method can improve the utilization rate of fingerprint information and the fingerprint retrieval efficiency.

The invention discloses a massive digital audio fingerprint storage and retrieval method, which comprises two steps of digital audio fingerprint storage and digital audio fingerprint retrieval, and the digital audio fingerprint storage comprises the following steps of: performing Hash processing on an audio file to be stored to obtain a unique identifier of each audio file; extracting audio fingerprints of the audio files to be stored, wherein the audio fingerprint of each audio file forms a character string; taking an identifier obtained after Hash of each to-be-processed audio file as a Rowkey of the Hbase, taking the fingerprint character string as a value of one column, taking meta information as a value of the other column, writing the values into a table of the Hbase, and writing thevalues into the Hbase; and taking identifier and fingerprint character string obtained after Hash of each audio file to be processed as two fields of document corresponding to audio file to be processed, and writing two fields into ES. Near-real-time full-text search and distributed characteristics of ES are utilized, and concurrent and real-time query is supported while high-efficiency query isguaranteed.

The invention discloses a video fingerprint extraction and retrieval method, which is used for extracting the texture information from an image sequence in a video, so as to obtain the marking information indicating the difference of the video from other videos. The video fingerprint extraction method comprises the following steps: the video is decoded to obtain YUV data of each image; the Y component of each image is extracted, the gradient information of the Y component is extracted through the Sobel operator, and the first 10 percent of points with the maximum gradient range in the horizontal (X) direction and the first 10 percent of points with the maximum gradient range in the vertical (Y) direction are respectively reserved; each video frame image is divided into N*N areas, the gradient distributions in the forward and the backward X directions as well as in the forward and the backward Y directions in each area are obtained through statistics, and the gradient distributions in all the areas are combined to form a normalized column diagram. The video fingerprint retrieval method comprises the following steps: a fingerprint sequence of a template video is slide on a fingerprint sequence of a video to be inquired about, and a cross correlation coefficient is calculated to find out a matching position with the maximum similarity. The invention can greatly simplify the video contents, and effectively improve the video retrieval efficiency.

The invention discloses a system and a method for retrieving contents of an audio / video. The method at least comprises the following steps of: receiving an audio / video segment; extracting the fingerprint characteristic of each frame of the audio / video segment; computing the anti-interference degree of a fingerprint of each frame by using an anti-interference degree computation model; sorting the frames according to the anti-interference degree of the fingerprint of each frame; and retrieving the fingerprints in a fingerprint database according to a frame sorting result. The fingerprint retrieval speed of the audio / video is greatly improved by computing the anti-interference degree of the fingerprint characteristic of each frame, sorting according to the anti-interference degree of the fingerprint characteristic of each frame and retrieving.

A retrieval method includes obtaining an audio sub-fingerprint to be retrieved, obtaining an audio sub-fingerprint to be retrieved, and retrieving the audio sub-fingerprint. According to the audio sub-fingerprint, reading an inverted item head address recorded in an index entry corresponding to the audio sub-fingerprint in a pre-generated address index table; Reading an inverted item entry corresponding to the audio sub-fingerprint in a pre-generated inverted table according to the inverted item head address, and obtaining each audio identifier ID recorded under the entry; According to the audio identifier ID, by a pre-generated audio identifier ID-Audio fingerprint database address mapping table, obtains the audio fingerprint file corresponding to each audio identifier ID in the audio fingerprint database storage address. The audio sub-fingerprint retrieval method provided by the present application draws on the inverted table method in text retrieval, and can quickly obtain the audioidentifier ID containing the audio sub-fingerprint. The present application also provides a retrieval method for identifying audio files and an audio inquiry system.

The invention relates to a parallel video fingerprint retrieval method which comprises the following steps of: (1) video fingerprint extraction step: carrying out video fingerprint coding on video files uploaded by users in a video database to extract video file fingerprints; (2) video file fingerprint packaging step: dividing the video file fingerprints into a plurality of video fingerprint packages according to a time sequence during uploading by the user, and distributing the plurality of video fingerprint packages to a plurality of computers calculating in parallel; (3) parallel retrieval step: carrying out parallel calculation retrievals in the computers with the video fingerprint packages within corresponding time periods according to the video fingerprints and appointed time periods of the video files to be retrieved; and (4) summarizing parallel calculation retrieval results. In the invention, the audio and video file fingerprints are packaged according to the time sequence and transmitted to the computers calculating in parallel to carry out the retrievals on the videos and the video fingerprints, thereby reducing unnecessary calculated amount and greatly improving the speed and the efficiency of fingerprint retrievals.

The invention discloses a posture correcting based fingerprint retrieval method and system. The retrieval method includes steps of acquiring posture information of a plurality of fingerprint images in a fingerprint map database; subjecting each fingerprint image to posture correcting; obtaining errors of the posture correcting algorithm by training the fingerprint images; partitioning each fingerprint image into many areas according to the errors and setting up a reverse index table, a fingerprint map database and a reverse index table for training the fingerprints in the same area in the many fingerprint images; studying the number of detail points corresponding to keywords and mapping among weights and establishing weighting functions; subjecting inquired fingerprints to posture correcting and retrieving via the weighting functions and the reverse index tables of the fingerprint map database. According to the fingerprint retrieval method in the embodiment, by subjecting the many fingerprint images in the fingerprint map database to posture correcting and partitioning each fingerprint image and by retrieving the fingerprint map database after subjecting the inquired fingerprints to posture correcting, retrieval accuracy and efficiency are improved.

The invention relates to a fingerprint feature expression technology and a fingerprint retrieval technology, and provides a fingerprint database quick retrieval method based on a core detail point support system. And the core detail point support system obtains each detail point of the fingerprint image one by one as a core detail point, the detail points except the core detail points are used assupport detail points, and the core detail points and the support detail points form the core detail point support system. And the support detail point feature vector is enabled to correspond to the core detail point and the fingerprint template ID by using a hash table. And during retrieval, the hash table is searched by using the support detail points, the ID of the core detail point with the highest occurrence frequency is counted, the ID frequency of the fingerprint template in the matched detail points is counted, and the most similar fingerprint template is found. According to the method, a two-dimensional feature vector is constructed, the size of a hash table is reduced, and the space complexity is reduced; the constructed feature vectors contain more fingerprint image information,the total number of the feature vectors converted from the fingerprint images is small, the hash table addressing frequency during retrieval is reduced, and the retrieval time complexity is reduced.

The invention is used for fingerprint indexing of an ultra-large type fingerprint database. The invention discloses an indexing structure and retrieval method for an ultra-large type fingerprint database, belongs to the computer identification field, and in particular relates to the fields such as fingerprint identification, database, data retrieval, image processing, etc. The invention disclosesa construction method for an ultra-large type fingerprint database retrieval system structure, an index evaluation method based on index matching error probability, a fingerprint indexing method based on layering depth index, and an index evaluation method based on one-edge and multiple-time matching structure characteristics for quickening the indexing of the ultra-large type fingerprint database, which have practical value and common applicability. In the off-line database establishment procedure, three searchers with complementary functions of hash indexing, layering depth indexing and characteristic value indexing are built according to the index evaluation method based on index matching error probability; and in the on-line indexing procedure, searchers according to the index evaluation method based on index matching error probability and the index evaluation method based on one-edge and multiple-time matching structure characteristics are built.

The invention provides a high-resolution fingerprint retrieval method based on sweat pores and multi-graph matching. The high-resolution fingerprint retrieval method sequentially comprises the following steps: a high-resolution fingerprint registration step: establishing and calculating a binary feature vector of registered sweat pores extracted from registered fingerprints of a registration library; a sweat pore retrieval step: establishing and calculating a binary feature vector of a test sweat pore extracted from the test fingerprint; a multi-graph matching step: extracting required nearestneighbor training sweat pores for each test sweat pore, and removing wrong nearest neighbor training sweat pores by adopting a multi-graph matching algorithm in combination with the spatial distribution relationship of the test sweat pores; and an output step: finally recommending a candidate fingerprint sequence according to a result of the sweat pore retrieval step. The high-resolution fingerprint retrieval method disclosed by the invention has the beneficial effects that retrieval is carried out through the sweat pore local characteristics in the fingerprint, and as sweat pores are richerthan other characteristics, the high-resolution fingerprint retrieval method can effectively improve the fingerprint retrieval precision.

The invention discloses a fingerprint retrieval method based on sweat pore features and a neural network. The method comprises the following steps: acquiring a fingerprint image; extracting global features of the fingerprint image and sweat pore features of the fingerprint image based on a trained feature extraction model; and screening and matching the fingerprint image for a plurality of times according to the global features of the fingerprint image and the sweat pore features of the fingerprint image to obtain a retrieval result of the fingerprint image. The global feature and the sweat pore feature of the fingerprint image are extracted, the fingerprint image is screened and matched for multiple times based on the two features, and the feature information can be effectively extracted by the feature extraction model due to rich sweat pore features, so that the fingerprint image retrieval speed and precision are improved.

The invention discloses a fast repetitive data block identification method based on 3D stacked memory. The method comprises the following steps: sending a data block fingerprint to a 3D stacked memory; Searching and storing the fingerprints of the data blocks in the 3D stack memory; returning fingerprint retrieval results to that CPU by the 3D stack memory. The 3D stack memory is used to store data block fingerprints, 3D stack memory is formed by stacking a plurality of DRAM chips and a logic layer chip, and is connected through TSV technology. That logic layer accesses the memory layer through TSV without passing through a data bus, the speed is fast, unnecessary data movement on the bus is avoided, and the time for accessing the memory is reduced; The invention classifies the data fingerprints, divides the memory chip, stores each class of fingerprints in one division, embeds a plurality of calculation units and a route in the logic layer, and routes the data fingerprints to each calculation unit, avoids the communication overhead between the calculation units, and reduces the energy consumption in the fingerprint search process.

Methods are provided for both single modal and multimodal fault diagnosis. In a method, a fault fingerprint is constructed based on a fault event using an invariant model. A similarity matrix between the fault fingerprint and one or more historical representative fingerprints are derived using dynamic time warping and at least one convolution. A feature vector in a feature subspace for the fault fingerprint is generated. The feature vector includes at least one status of at least one system component during the fault event. A corrective action correlated to the fault fingerprint is determined. The corrective action is initiated on a hardware device to mitigate expected harm to at least one item selected from the group consisting of the hardware device, another hardware device related to the hardware device, and a person related to the hardware device.

The invention discloses an audio fingerprint retrieval library updating method and device, equipment and a medium, and the method comprises the steps: obtaining a target song ID from a to-be-extracted feature queue when it is monitored that the to-be-extracted feature queue is non-empty; performing feature extraction on the to-be-processed audio data corresponding to the target song ID to obtain target audio feature data; storing the target song ID into a to-be-inserted fingerprint queue, and storing the target audio feature data into a preset storage area according to the target song ID; when it is monitored that the to-be-inserted fingerprint queue is non-empty, obtaining an earliest stored target song ID from the to-be-inserted fingerprint queue; encoding target audio feature data corresponding to the earliest stored target song ID to obtain a corresponding target audio fingerprint; and sending the target audio fingerprint to a target online server, so that the target online server inserts the target audio fingerprint into an audio fingerprint retrieval library. In this way, the audio fingerprint retrieval library can be updated in real time.

The invention discloses a device for automatically distributing materials retrieved by fingerprints, which relates to the field of mechanical equipment. It mainly solves the problems of short time for picking up materials and wrong materials during work. The main structure of the device includes a storage table, so The storage platform is equipped with information input mechanism, identification mechanism, self-service retrieving mechanism, and material swing mechanism. The swing mechanism is located on the side of the self-retrieving mechanism. Put it in the groove, through the fingerprint identification of the staff, the controller automatically selects the material storage box at the corresponding position and takes it out, and moves it into the picking channel under the action of gravity, which improves the work efficiency. The invention has the advantages of simple structure and strong practicability.

The invention discloses an audio fingerprint retrieval method and a retrieval device. The audio fingerprint retrieval method includes the steps of carrying out framing process to audio signals of an audio clip, determining key frames in each subframe obtained through the framing processing by means of an onset detection function (ODF) method, obtaining audio fingerprints of the key frames, carrying out retrieval according to the obtained audio fingerprints, and obtaining audio file information corresponding to the audio clip. Correspondingly, the invention further discloses an audio fingerprint storage method and a storage device. The audio fingerprint storage method and the storage device can greatly reduce wasted memory due to audio retrieval, greatly reduce wasted time due to the audio retrieval, and improve retrieval efficiency on the premise of ensuring accuracy rate of audio fingerprint retrieval.

A code fingerprint anti-counterfeiting method provided by the present invention includes establishing a code fingerprint database, establishing a code fingerprint of a product to be tested, and retrieving and verifying; establishing a code fingerprint database includes extracting local feature points of a mark area image; Local feature points, the remaining local feature points form a significant local feature point map; obtain block detailed feature vectors according to the significant local feature point map; establish key area feature codes based on the significant local feature point map and block detailed feature vectors fingerprint. The above code fingerprint anti-counterfeiting method collects the image of the commodity logo area on the optical device to contain the difference in printing characteristics, and after processing it through a certain method, a uniquely identifiable code fingerprint is established and stored in the code fingerprint database. By collecting the corresponding image of the commodity to be tested The image of the marked area can be used to determine whether it has a unique code fingerprint in the code fingerprint library. The code fingerprint of this code fingerprint anti-counterfeiting method cannot be deciphered and copied, which can prevent counterfeiters from being prevented.

The invention discloses an audio fingerprint retrieval method, device and equipment. The method comprises the following steps that: acquiring audio information input into a retrieval engine of an audio fingerprint platform, extracting audio features of the audio information, retrieving audio fingerprint retrieval results of a plurality of audio fingerprints matched with the audio features on the audio fingerprint platform, and displaying the audio fingerprint retrieval results of the plurality of audio fingerprints according to a correlation degree sequence. In this way, the audio fingerprintretrieval results of the multiple audio fingerprints associated with the audio features corresponding to the audio information input into the retrieval engine of the audio fingerprint platform can beobtained through retrieval.

The invention discloses a system and a method for retrieving contents of an audio / video. The method at least comprises the following steps of: receiving an audio / video segment; extracting the fingerprint characteristic of each frame of the audio / video segment; computing the anti-interference degree of a fingerprint of each frame by using an anti-interference degree computation model; sorting the frames according to the anti-interference degree of the fingerprint of each frame; and retrieving the fingerprints in a fingerprint database according to a frame sorting result. The fingerprint retrieval speed of the audio / video is greatly improved by computing the anti-interference degree of the fingerprint characteristic of each frame, sorting according to the anti-interference degree of the fingerprint characteristic of each frame and retrieving.

The invention relates to a parallel video fingerprint retrieval method which comprises the following steps of: (1) video fingerprint extraction step: carrying out video fingerprint coding on video files uploaded by users in a video database to extract video file fingerprints; (2) video file fingerprint packaging step: dividing the video file fingerprints into a plurality of video fingerprint packages according to a time sequence during uploading by the user, and distributing the plurality of video fingerprint packages to a plurality of computers calculating in parallel; (3) parallel retrieval step: carrying out parallel calculation retrievals in the computers with the video fingerprint packages within corresponding time periods according to the video fingerprints and appointed time periods of the video files to be retrieved; and (4) summarizing parallel calculation retrieval results. In the invention, the audio and video file fingerprints are packaged according to the time sequence and transmitted to the computers calculating in parallel to carry out the retrievals on the videos and the video fingerprints, thereby reducing unnecessary calculated amount and greatly improving the speed and the efficiency of fingerprint retrievals.

The invention relates to a music and human voice separation method based on a U-shaped network and an audio fingerprint, and the method comprises the steps: obtaining a separated accompaniment audio through a trained music separation model when a mixed audio is inputted; then extracting fingerprints in the accompaniment audio by adopting a fingerprint extraction algorithm; then sending the fingerprints into an audio fingerprint database, and retrieving corresponding matched songs and time points offset; then, adjusting the volume of the retrieved song; matching sampling points within a period of time before and after the song matching time point offset to obtain a section of audio waveform; accurately matching the separated audio with the volume of the waveform; and finally, subtracting the waveform from the mixed audio to finally obtain a relatively pure audio. According to the method, a music separation model based on a U-shaped network architecture is designed, and audio fingerprint retrieval is added behind the model for re-separation, so that the separation effect difference is obviously reduced, and pure human voice is obtained.

The invention discloses a video fingerprint extraction and retrieval method, which is used for extracting the texture information from an image sequence in a video, so as to obtain the marking information indicating the difference of the video from other videos. The video fingerprint extraction method comprises the following steps: the video is decoded to obtain YUV data of each image; the Y component of each image is extracted, the gradient information of the Y component is extracted through the Sobel operator, and the first 10 percent of points with the maximum gradient range in the horizontal (X) direction and the first 10 percent of points with the maximum gradient range in the vertical (Y) direction are respectively reserved; each video frame image is divided into N*N areas, the gradient distributions in the forward and the backward X directions as well as in the forward and the backward Y directions in each area are obtained through statistics, and the gradient distributions in all the areas are combined to form a normalized column diagram. The video fingerprint retrieval method comprises the following steps: a fingerprint sequence of a template video is slide on a fingerprint sequence of a video to be inquired about, and a cross correlation coefficient is calculated to find out a matching position with the maximum similarity. The invention can greatly simplify the video contents, and effectively improve the video retrieval efficiency.

The present invention provides a fingerprint retrieval method based on global direction information, comprising: A, establishing a fingerprint global direction information feature representation model; B, extracting the global direction information of the fingerprint database and the input fingerprint image, using the global direction information The feature representation model constructs feature vectors, which are then retrieved. The invention can improve the utilization rate of fingerprint information and improve the efficiency of fingerprint retrieval.

The present invention involves a method and human voice separation method based on U -shaped networks and audio fingerprints: when entering hybrid audio, first through the trained music separation model to obtain the separation accompaniment audio; then use fingerprint extraction algorithm to extract it out to extract out the fingerprint extraction algorithm to extract it out to extract out and extract it.Fingerprints in the accompaniment audio; then send these fingerprints into the audio fingerprint database to retrieve the corresponding matching songs and time points of offset; then adjust the volume of the retrieved songs;The sampling point is matched to get a period of audio waveform; then accurately match the separation audio with the volume of this waveform; finally mixed audio minus this waveform and finally get a more pure audio.This method designs a U -shaped network architecture -based music separation model, and adding audio fingerprint retrieval after the model to separate the separation of the separation effect, which significantly reduces the gap between the separation effect and obtains pure human voice.

The invention discloses a video fingerprint retrieval method based on coarse and fine granularity, which searches and compares video fingerprints in a layered and granular retrieval manner. Use the local sensitive hash algorithm to hash all the video fingerprints in the video fingerprint database and the video fingerprints to be retrieved, so as to find the video fingerprints most similar to the video fingerprints to be retrieved, complete the coarse-grained search, and effectively reduce the cost of video retrieval The time overhead is well prepared; in the case of fine-grained search, through the improved fast video fingerprint matching mechanism based on the biological sequence comparison technology BLAST, it can quickly find out whether the video fingerprint to be retrieved exists in the video fingerprint database, and can also perform video clip matching. Matching search, and according to the characteristics of video fingerprints, determine the complete video to which the video segment to be retrieved belongs and the specific time position in which it appears, thus ensuring the accuracy and real-time performance of video search.