47results about How to "High precision extraction" patented technology

The invention discloses a cold atom beam interference gyro device, which comprises a cold atom beam system, a coherent atom beam operating system, and a gyro rotating signal extracting system. The cold atom beam system continuously emits | 1 > state cold atom beams as a matter wave source; the coherent atom beam operating system is composed of three gratings placed on the emitting path of | 1 > state cold atom beams in turn, and three phase modulators for modulating the phases of three gratings, respectively; | 1 > state cold atom beams orderly enter the three grating for splitting, reflecting and splitting again so as to obtain | 2 > energy state atom beam and | 1 > state cold atom beam interference signals; the | 2 > energy state atom beam interference signals are induced to produce fluorescent signals by the probing laser of the gyro rotating signal extracting system, a photoelectric detecting device detects the fluorescent signals and transmits the signals to a computer for processing to obtain gyro rotating angular-speed signals. With this device disclosed by the invention, high-precision gyro absolute rotating signals relative to inertial space and with high requirement on environment can be obtained.

The invention relates to a multi-line lidar and camera joint calibration method based on fine radar scanning edge points, which mainly relates to the technical fields of robot vision, multi-sensor fusion and the like. Because of the influence of the resolution of lidar, the scanned edge points are often not accurate enough, so the calibration results are not accurate enough. According to the characteristic that the range of the lidar point is abrupt at the edge, the invention searches and compares for many times, and takes the point closer to the edge as the standard point. By detecting the circle in the camera image and radar edge points, the translation between the camera and the lidar is calculated according to the pinhole camera model. The calibration parameter C is searched in the neighborhood space of the obtained translation vector to find the calibration result that minimizes the projection error. The invention can extract the points swept by the lidar on the edge of the objectwith high precision, avoids the problem of insufficient precision caused by low resolution of the lidar, and improves the calibration precision.

The invention provides a shoreline extraction method of a remote sensing image based on an object-oriented method and a membership rule thereof. The method obtains an optimal segmentation scale by evaluating multi-scale segmentation and segments an image based on homogeneity principle. Then, the feature samples are constructed to obtain the features of the underlying surface. Secondly, attributesare assigned to the image object unit based on membership rules. Finally, the land information is divided into water and non-water binary information images, and the coastline information is obtainedby line reduction algorithm after grating loss conversion. The method of the invention realizes the high-precision extraction of the coastline, has good fitting degree in most areas, and can be betterapplied to the extraction of the coastline in a large range.

A continuous-region-extraction unit extracts a continuous-region having a voxel value corresponding to an air-region in bronchi from a three-dimensional-medical-image. A tree-structure-construction unit constructs a tree-structure corresponding to the continuous-region. A linear-structure-extraction unit extracts plural linear-structures representing fragments of small-bronchi by analyzing a local density structure in a neighborhood of each point in the three-dimensional-medical-image. A tree-structure-reconstruction unit reconstructs a graph-structure representing the whole bronchi by connecting a node constituting the graph-structure of large-bronchi and a node representing the linear-structures of small-bronchi. At this time, different cost functions are used for a segment connecting the node of the large-bronchi and the node representing the linear-structures of the small-bronchi and a segment connecting the nodes representing the linear-structures to each other. The cost function in the former segment is defined in such a manner that the segment is more likely to be connected as a change in voxel values is smaller.

The invention provides an image processing program and an image processing device, recorded words, prints, marks and the like are extracted in a high accuracy from document images comprising color bias or fuzzy color. The image processing method of the invention is composed of: a background removing and data generating part for removing background parts from the color images or bright images, generating, and generating the background removing data which represents the parts other than the background; a profile color converting and data generating part for, in the parts other than the background in the color images or bright images, generating the data which converts the profile color of the parts other than the background into pixel color at inner side of the profile of the parts other than the background.

The invention, which belongs to the technical field of electrical variable measurement, provides an adaptive-notch-filter-based harmonic wave extraction method. The method comprises: a, collecting a to-be-measured signal by a sampling device to obtain an input signal u(t); and b, carrying out harmonic wave extraction on the input signal u(t) so as to obtain each harmonic instantaneous value . / xh, a fundamental frequency omega 1, a direct current deviation A0, and each harmonic amplitude Ah. With the method, a problem of measurement error occurrence when direct current deviation occurs in the input and the amplitude is changed according to the original adaptive notch filter (ANF) method can be solved. Even though the amplitude and frequency of the input signal are changed suddenly and even the direct-current deviation is added, the fundamental frequency, the direct current deviation, and each harmonic instantaneous value and each harmonic amplitude can be extracted in real time with high precision based on the algorithm.

The invention relates to a double-field variable-focus three-dimensional measurement system. The system comprises a primary-secondary mirror imaging measuring system, a two-dimensional tracking rotating-mirror device and an adaptive control system. The primary-secondary mirror imaging measuring system is composed of a telephoto subsystem, a short-focus subsystem and a launching light source. The telephoto is composed of a main lens, a first spectroscope, a second spectroscope and a detector. The short-focus system is composed of the main lens, the first spectroscope, the second spectroscope and the detector. The two-dimensional tracking rotating-mirror device is composed of a reflector, a electric rotating motor, a regulating handle, an upper half shaft, a lower half shaft and a horizontal shaft. The adaptive control system is composed of a computer, an input and output (I / O) control circuit, an electromotor and a signal processing and controlling circuit. The system has the advantages of being tight in arrangement, high in measuring efficiency, convenient to measure and is suitable for the three-dimensional measuring of moving and static objects.

The invention relates to a joint calibration method based on a laser radar and a camera and a computer readable storage medium. The joint calibration method comprises the following steps of: carryingout edge detection on a laser radar point cloud extracted by the laser radar, extracting edge points of a background part, searching a foreground edge point corresponding to each background edge pointthrough a depth difference of laser radar points; performing repeated searching and comparison to obtain foreground edge points closer to an edge, then respectively extracting the laser radar edge point cloud and the circle center and radius of a circle in an edge image acquired by the camera at a same moment, calculating a translation vector between the camera and the laser radar through a pinhole camera model, searching calibration parameters in a neighborhood space of the obtained translation vector so as to find a calibration result minimizing the projection error. Points scanned by the laser radar at the edge of an object can be extracted with high precision, and the problem of insufficient precision caused by low resolution of the laser radar can be avoided. Through the obtained precise edge points, a more accurate calibration result can be obtained.

The invention relates to an object detecting method for detecting a specific kind of object such as a human head and a human face from an image expressed by two-dimensionally arrayed pixels, the object detecting method including an image group producing step of producing an image group including an original image of the object detecting target and at least one thinned-out image by thinning out pixels constituting the original image at a predetermined rate or by thinning out the pixels at the predetermined rate in a stepwise manner; and a stepwise detection step of detecting the specific kind of object from the original image by sequentially repeating plural extraction processes from an extraction process of applying a filter acting on a relatively small region to a relatively small image toward an extraction process of applying a filter acting on a relatively wide region to a relatively large image.

A layout analysis program, a layout analysis apparatus, layout analysis method and a medium can highly accurately extract a text block from an image if the image is a color image. The layout analysis program causes a computer to execute a divided region extracting step that extracts a region partitioned by a pattern according to a binary image so as to use the outcome of extraction as divided region, a set of character elements extracting step that extracts a set of the character elements extracted by a first binary image layout analysis process for each extracted divided region so as to use the outcome of extraction as set of character elements, a text block extracting step that extracts a region including the extracted set of character elements in each divided region so as to avoid overlapping the non-character elements extracted by a second binary image layout analysis process and use the outcome of extraction as text block and a layout information generating step that generates layout information according to the text block and the non-character elements extracted by the second binary image layout analysis process.

A reproduction equipment converts an identification code of the reproduction equipment into a plurality of signal sequences using a correspondence table and embeds the obtained plurality of signal sequences in AV signals as a digital watermark. A reproduction equipment specifying equipment compares correlation values of the AV signals, containing the digital watermark embedded therein with all signal sequences contained in a correspondence table. A comparison of the correlation values with a previously set threshold value determines the embedded signal sequences and the identification code of the reproduction equipment which may be extracted.

The invention discloses an image processing method and device for a mining area. The method comprises the following steps: inversing and interpreting a first multiband image in remote sensing images to obtain a land status coverage set which comprises a vegetation area coverage and a non-vegetation area coverage; performing raster calculation to a second multiband image in the remote sensing images to obtain a vegetation coverage diagram; cutting images to the vegetation coverage diagram by the vegetation area coverage to obtain a vegetation coverage; and synthesizing the vegetation area coverage and the non-vegetation area coverage to obtain an ecological diagram. Through the invention, ecological lands in the mining area are identified with high precision, and the ecological status of the mining area is comprehensively and completely reflected.

The invention provides an extraction device and an extraction method of an area of interest. The area of interest is extracted from an image at a high precision, and the interest degree of the area of interest is calculated. The extraction device of the area of interest comprises an extraction component, a retrieval component and an interest degree determining component, wherein the extraction component extracts one or a plurality of partial areas from an input image; the retrieval component retrieves an image similar to the partial area from an image database which stores a plurality of images in regard to the respective partial area extracted by the extraction component; and the interest degree determining component determines the interest degree of the respective partial area on the basis of the retrieval result of the retrieval component.

The invention discloses a signal detection method and circuit for a high precision optics voltage transformer and belongs to the technology field of an optics voltage transformer. The signal detection method and circuit for the high precision optics voltage transformer utilizes a band-pass filter of a narrow-band to replace the band-pass filter of a wide-band, utilizes a digital low pass filter and a control unit to replace a primary signal processing unit; and utilizes a robust control algorithm to obtain index stability during the signals detection process to improve the optics voltage transformer dynamic property which varies with the luminous intensity. The signal detection method and circuit for the high precision optics voltage transformer restrains white noises very well to obtain a closed loop error signal of high precision. By utilizing the detection method and circuit, the optics voltage transformer obtains + / - 0.2% precision, 24.5 kilohertz wide-band and 25.2 microseconds speediness step response which satisfy the application of electricity measuring and electricity protection fields.

According to the present invention, a brightness change (dI) is calculated (21) with respect to each pixel, using a plurality of captured images obtained by imaging performed under different lighting conditions; a texture change (dF) is calculated (22) with respect to each pixel, using a plurality of captured images obtained by imaging performed at different times; and a subject region is extracted (23) on the basis of the brightness change (dI) and the texture change (dF). A change, between the captured images, in texture feature (F) for each pixel of the images is calculated as the texture change (dF). A subject can thus be extracted with high accuracy even when there is a change in ambient light or background.

The invention relates to a restoration effect prediction modeling method of an organic polluted underground water in-situ chemical oxidation process. The method specifically comprises the following steps: step 1, performing detailed investigation on the basis of a polluted site; 2, determining an in-situ chemical oxidation process of the polluted site, and determining a related parameter range; 3, using a Plackett-Burman test for determining main influence factors of the effect of the repairing technology; 4, analyzing the response of different independent parameters by using Design Expert software to obtain test combinations under different test conditions; and 5, carrying out a laboratory small-scale test based on the simulated sample, and the like. The method is simpler and more convenient, the requirement for mathematical knowledge of operators is low, data is used as a support, the statistics principle is used as a support, and the problems of model uncertainty and mathematical calculation difficulty caused by the underground complex environment are emphatically solved.

Character string heights and line spacing values are acquired as character string information on a document image, and fluctuations in the character string height and fluctuations in the line spacing value are calculated as variances. If the calculated variance is equal to or lower than a threshold value, the character string information is determined as being appropriate for use in extracting line-spacing watermarks, and line-spacing watermark information is extracted from the character string information.

The invention relates to an unsupervised change detection device based on a full polarization sequential SAR image; the device comprises the following units: an image preprocessing unit used for registering and filtering a pair of full polarization sequential SAR images of different time phases in a same geological scope; a preliminary difference image forming unit used for carrying out hypothesis examination for corresponding positions in the pair of full polarization sequential SAR images, thus forming a preliminary difference image; an optimal threshold select unit using a KI algorithm to select the optimal threshold of the formed preliminary difference image; a detection result forming unit using the selected optimal threshold to carry out binary segmentation for the preliminary difference image, endowing values representing no changes for pixels smaller than the threshold, endowing values representing changes for pixels bigger than the threshold, and finally forming the detection result. The method and device can fully utilize the SAR image full polarization information, thus automatically extracting change information of SAR images of different time phases with high precision.

The invention provides a coastal zone wetland ecological safety pattern optimization method and device based on machine learning, and the method comprises the steps: employing a support vector machineand a random forest method to carry out the supervised classification extraction of a pixel scale of a coastal zone wetland of a target coastal zone, so as to obtain a coastal zone wetland extractionresult; performing shape recognition on each wetland patch in the coastal zone wetland extraction result by adopting a patch shape recognition algorithm based on an information theory; and comparingecological disturbance degree differences of wetland patches with different shapes by utilizing variance analysis to obtain an optimal coastal zone wetland landscape type with statistical significance, thereby obtaining a landscape ecological safety pattern mode capable of maximizing ecological toughness. The method and deice can assist in guiding coastal zone wetland landscape pattern optimization, and landscape planning thought and theoretical basis are provided.

The invention provides a connector holder, an assembled drawing board and a spring component. The connector holder can correspond to the connectors having various widths and depths with fewer tools, by correctly arranging the tools relative to the connectors on a wiring plate dispensing with time and assembling skills. The connector holder of the invention comprises a holder main body, a spring component supporter fixed on the holder main body and the spring component mounted on the spring component supporter, wherein the holder main body comprises a comb toothed connector supporter and a connecting part connected to the spring component supporter; the spring component supporter comprises a connected part connected to the holder main body and a rear panel; the spring component comprises a spring part which is freely detachably mounted on the rear panel for supplying an elastic force on the direction of the comb toothed connector supporter; and a connector receiving part is formed between the comb toothed connector supporter and the spring part.

The invention discloses a method, a device and an image retrieval method for extracting a hash code from an image. The device comprises the following steps: a hash code extraction model being constructed, wherein the model comprises an encoder and a decoder, and the encoder is composed of a multi-layer depth neural network DNN; extracting hash code from image data and outputting it to decoder, wherein the decoder is composed of multi-layer DNN, which converts the input hash code into image; regularizing the output of the last layer of the decoder so as to ensure that the output of the hidden layer of DNN is close to the hash code as far as possible, thereby simplifying the network structure of the decoder and forcing the encoder to extract high-quality hash code to obtain an anti-redundanthash code depth extraction model; training the depth extraction model of anti-redundant hash codes to determine the parameters in the model; the hash code being extracted from the image by using thetrained anti-redundant hash code depth extraction model encoder. The method can effectively reduce the redundancy of the encoded spatial information and extract the image hash code with high precisionby using all dimensions.

The invention provides a signal processing device, including: an extraction section that extracts, from an input digital signal, a decoding target signal at an extraction timing that has been determined as a timing for extracting the decoding target signal; a decoding section that decodes the decoding target signal by estimating, by a maximum likelihood decoding, a candidate for a decoding result of the decoding target signal extracted by the extraction section and detecting a maximum likelihood decoding result; and an adjustment section that adjusts the extraction timing using a likelihood of the candidate for the decoding result estimated by the decoding section.

The invention discloses a method and device for extracting a hash code from an image, and an image retrieval method and device. The method includes: constructing a hash code extraction model, the model includes an encoder and a decoder; wherein, the encoder consists of multiple The layer deep neural network DNN is composed, and the hash code is extracted from the image data and output to the decoder. The decoder is composed of multiple layers of DNN, and the input hash code is converted into an image; in the encoder, the hidden layer coding redundancy is measured. Regularize the hidden layer coding to reduce the redundancy of coding space information, and obtain the anti-redundancy hash code depth extraction model; train the anti-redundancy hash code depth extraction model to determine the parameters in the model; use the trained anti-redundancy hash code depth extraction model Encoder in Redundant Hash Code Deep Extraction Model to extract hash codes from images. This method can effectively reduce the redundancy of coding space information, effectively utilize all dimensions, extract image hash codes with high precision, and effectively improve the accuracy of related application fields such as image retrieval.

Provided is a pallet sensing device or the like which can accurately extract a straight line corresponding to the front surface of a pallet without moving a two-dimensional distance measuring device in the vertical direction, and can sense the position and posture of the pallet in a short time. The pallet sensing device is provided with: a point group acquisition unit configured so as to acquire point group data in which point groups measured by the two-dimensional distance measurement device are represented on a depth map; a straight line detection unit configured to detect a straight line corresponding to the front surface of the pallet on the basis of a point group in a region deduced to include the front surface of the pallet in the point group data; a line segment detection unit configured to detect a line segment indicating a front surface of the pallet on the basis of the straight line; and a position and posture acquisition unit configured to acquire the position and posture of the pallet on the basis of the line segment. The straight line detection unit acquires one or more straight line candidates as candidates for straight lines corresponding to the front of the pallet, and assigns, to each of the one or more straight line candidates, a score having a lower priority than a constant of other point groups when selected to a point group on the front side that is a predetermined distance or more from the straight line candidate. A straight line corresponding to the front of the pallet is selected from among the one or more straight line candidates on the basis of a score accumulation value obtained by accumulating scores for the point groups in the region deduced to include the front of the pallet.