888 results about "High complexity" patented technology

The invention belongs to the technical field of digital image processing and mode identification, and specifically relates to an FPGA-based deep convolution neural network realizing method. The hardware platform for realizing the method is XilinxZYNQ-7030 programmable sheet SoC, and an FPGA and an ARM Cortex A9 processor are built in the hardware platform. Trained network model parameters are loaded to an FPGA end, pretreatment for input data is conducted at an ARM end, and the result is transmitted to the FPGA end. Convolution calculation and down-sampling of a deep convolution neural network are realized at the FPGA end to form data characteristic vectors and transmit the data characteristic vectors to the ARM end, thus completing characteristic classification calculation. Rapid parallel processing and extremely low-power high-performance calculation characteristics of FPGA are utilized to realize convolution calculation which has the highest complexity in a deep convolution neural network model. The algorithm efficiency is greatly improved, and the power consumption is reduced while ensuring algorithm correct rate.

The present invention provides methods and compositions directed toward assays of a broad range of analytes using specific targeting chemicals that bind to the analytes. The assays are founded on the use of coded assay articles to which the targeting chemicals are attached. Additionally the codes are such that they are interrogated and determined in real time. The target is analyzed as to identity, presence and quantity in real time. The methods and compositions of the invention are highly suitable for use in high-complexity multiplexed assay systems. All the methods and compositions are based on assay article that includes an optical substrate to which the chemical is bound, and in which is disposed at least one diffraction grating. The grating provides an output optical signal when illuminated by an incident light signal which is indicative of the code in the substrate. In general, coded assay article or sets thereof are employed in assay methods, including multiplexed assay methods, according to which a sample is contacted with an article or a set, and any analytes that bind to the attached chemical are identified according to the code, detected and/or quantitated.

The invention discloses an MIMO system and an application method thereof for adaptive antenna selection. The MIMO system comprises a transmitter, a receiver and a wireless channel; the process of the antenna selection which is realized in the MIMO system adopts an iterative feedback structure; a subset is selected quickly from a large amount of antenna sets by using the characteristic of 60 GHz channel without accurate channel estimation, so that a subchannel corresponding to an antenna subarray is optimal in the sense of a defined target function; and high speed data transmission is performed by using the selected optimal antenna subarray. The invention provides the rapid and efficient application method for the MIMO system along with low cost, and solves the technical problems of high complexity and low efficiency of the traditional antenna selection method in the large-scale the MIMO system. An experiment proves that: the adopted method can obtain good performance only through several iterations and low computational complexity; and the system can better support the application of the large-scale the antenna array.

A method of encoding data for transmissions from a source to a destination over a communications channel is provided. The method operates on an ordered set of source symbols and may generate zero or more redundant symbols from the source symbols, wherein data is encoded in a first step according to a simple FEC code and in a second step, data is encoded according to a second FEC code, more complex than the first FEC code. The first FEC code and/or the second FEC code might comprise coding known in the art. These steps result in two groups of encoded data in such a way that a low-complexity receiver may make use of one of the groups of encoded data while higher complexity receivers may make use of both groups of encoded data.

An apparatus, method and system for geographical tracking entry and/or exiting of an asset into and/or out of a defined geographical boundary and reporting the same. Entry and exit tests compare position fixes with various thresholds and parameters to determine if the asset has entered or exited the geographical boundary. Tests are sequenced such that tests having lower levels of complexity (lower order) are performed before tests having higher levels of complexity (higher order). In this way, most position fixes are processed using computations having a lower order of mathematical complexity than conventionally implemented.

The invention relates to a general-purpose automatic test method for a locomotive switching power supply, and general-purpose automatic test system, and comprises a test control system, a programmable power supply unit, a programmable DC load unit and a programmable data acquisition unit, wherein the programmable power supply unit is connected with the test control system through a GPIB bus and provides excitation for the tested switching power supply; the programmable DC load unit is connected with the test control system through the GPIB bus and provides a load for the switching power supply to be tested; and the programmable data acquisition unit is connected with the test control system through the GPIB bus and acquires responses of the tested switching power supply. The output voltage, voltage regulation factor, load regulation factor and the like are tested by logging in virtual instrument software on a personal computer. The invention can well overcome the defects of insufficient compatibility, high complexity and low reliability of the test system in the prior art, and has the advantages of automation, general purpose, convenient operation and reliable performance.

The digital video referencing frame image is compressed block by block by applying lossless compression algorithm to pixel components with full length, or 1 bit, 2 bits, 3 bits or 4 bits LSB bits truncation. If a sub-block has high complexity which results in more than 3 bits error for most pixel components, a transfer algorithm with quantization and VLC coding is applied to compress this sub-block. Should the complexity is higher than a threshold or at least one sub-block having error of more than 3 bits for most pixel components, truncating 1 LSB bit of sub-block with simple pattern to save more bits to be allocate to code the sub-block with highest complex pattern.

Receivers accommodating frequency band selection methods in wireless communications systems are described. Different frequency bands are associated with different alternative carrier frequencies and/or base station cell and/or sector transmitter connection alternatives. Mobile node receivers include two receiver chains, each chain processing signals corresponding to a carrier. In some embodiments, each receiver chain includes its own controllable RF module, and individual carrier band selection is performed in each RF module. In some embodiments, the two receiver chains share a common RF module; however, each chain includes its own controllable baseband filter. In various embodiments, the first chain has higher complexity than the second chain. In most embodiments, each chain uses the same technology, e.g., spread spectrum OFDM or CDMA. Each chain obtains a quality indicator value on a different band and a comparison of quality indicator values is used in selecting the channel and carrier band for downlink traffic signaling.

The embodiment of the invention discloses a method, a system and a device for estimating time difference of arrival TDOA. The method for estimating the TDOA comprises the following steps that: time of arrival of data clustering estimated by a positioning reference point is received, and the time of arrival of the data clustering is detected at regular time by the positioning reference point according to a system precursor symbol, and is acquired by carrying out timing error estimation on an orthogonal frequency division multiplexing OFDM symbol containing a service LBS sequence based on positioning; and the TDOA is calculated according to the time of arrival of the data clustering estimated by different positioning reference points. Through the embodiment of the invention, the positioning reference point estimates the time of arrival of the data clustering, and a base station calculates the TDOA according to the time of arrival of the data clustering estimated by different positioning reference points so as to acquire specific position coordinates of a mobile station, thereby solving the problems of low accuracy and high complexity of the prior positioning algorithm.

The invention discloses a content awareness deep learning network-based remote sensing image super-resolution reconstruction method. A comprehensive measurement index and a calculation method for content complexity of images are proposed; based on this, the sample images are classified by the content complexity; deep GAN models with low, medium and high complexity are built and trained; and according to the content complexity of the to-be-classified super-resolution input images, corresponding networks are selected for performing reconstruction. In order to improve learning performance of a GAN, an optimized loss function definition is given. The method overcomes the over-fitting and under-fitting contradiction ubiquitous in machine learning-based super-resolution reconstruction, and effectively improves the super-resolution reconstruction precision of the remote sensing images.

An exemplary method includes receiving source code having a plurality of code segments, providing a desired level of quality for the source code, analyzing the source code to assign a complexity measure to each of the plurality of code segments and assigning a level of code coverage to each of the plurality of code segments based at least in part on the desired level of quality and the complexity measures. An exemplary system prioritizes quality improvements to source code based, in part, on a quality assessment. Such a system can improve code quality by assigning higher test coverage levels to modules with higher complexity.

A rapid target-identification and positioning system and a method relate to DSP and CPLD, FPGA technologies, as well as to image processing and mode recognition technologies. The system comprises a power unit, an imaging subsystem, a main processor, a memory subsystem, a control subsystem and an external interface circuit. The system is characterized in that being the calculating core of the whole system, the main processor is connected with all subsystems; and the imaging subsystem and the memory subsystem are respectively connected with the control subsystem; the digital signal collected, converted and input by the collecting unit of the image subsystem is submitted into the main processor and displayed by the display unit of the image subsystem after being processed; the power unit which is connected with the imaging subsystem, the main processor, the memory subsystem and the control subsystem can be used for supplying working voltage for collecting, displaying and processing signal and for the control of processing; the system has simple structure, high complexity, small size, high processing speed, and high identification rate.

The invention discloses a dynamic gesture identification method in an interactive system, comprising the following four steps of: acquiring an image, and obtaining a current image frame after denoising; obtaining a moving region of the image by calculating the difference of the current image frame and a reference image frame, and obtaining a binary image by a skin color detection method; carrying out gesture segmentation on the binary image by a gray level projection method to obtain a gesture region and a barycentric position; and obtaining a final identification result by a classification identification method according to the barycentric position and fingertip characteristics. The method can be used for overcoming the problems of high complexity, low real-time quality and low identification rate in traditional algorithms, has the advantages of simplicity in implementation, good real-time quality and high identification rate, and can be applied to the dynamic gesture identification very well. Moreover, the method can tolerate the translation of a definition gesture, the scale and the transformation of a rotation angle, and has a good robustness.

The present invention provides a method of detecting nucleotide sequence differences between two nucleic acid samples. The method employs a comparative genomic hybridization (CGH) technique to analyze the sequence differences between the samples. This method permits the identification of small sequence differences (e.g., sequence divergence of 1% or less) in nucleic acid samples of high complexity (e.g., an entire genome).

Methods and apparatus are disclosed for applying successive multi-rank beamforming strategies (e.g., successive preceding strategies) for the design of precoders over a set of parallel channels. Successive beamforming is applied to a narrow band channel model and is also applied for finer quantization of a single beamforming vector (e.g., recursive beamforming). A first embodiment provides the optimal approach with high complexity. An alternative embodiment provides successive beamforming for near optimal preceding selection with medium complexity. A low complexity method for precoder selection is also provided wherein a channel representative matrix for the set of parallel channels is determined and successive beamforming on the calculated channel representative is applied.

The invention provides a satellite-machine combined frequency domain imaging method of Bistatic Synthetic Aperture Radar (Bistatic SAR), which utilizes characteristics of the SA-BSAR system to obtain the analytical expression of two-dimensional spectrum of the system, and then utilizes Omega-k algorithm thought to obtain SA-BSAR two-dimensional Stolt mapping relation according to the expression of the two-dimensional spectrum; finally, two-dimensional Stolt interpolation can be approximately achieved by using inverse scale Fourier transform and phase multiplication in frequency domain through studying the two-dimensional Stolt mapping relation, so as to complete the compensation for the two-dimensional space variance of the system; the method can avoid the problem of huge calculation and high complexity caused by two-dimensional Stolt interpolation, thereby effectively achieving SA-BSAR imaging with high resolution.

The invention provides a manufacturing method of a high-speed PCB and the PCB. The manufacturing method includes the following steps that 1, lamination is carried out after a normal inner layer graph is manufactured; 2, copper reduction is carried out to achieve the size of 6 micrometers to 8 micrometers after lamination is completed; 3, +4.5 mil copper etching windowing is carried out on a deep micropore region and a first-order blind hole region overall; 4, laser drilling is carried out to complete deep micropore and first-order blind hole machining; 5, through holes needing resin hole plugging are mechanically drilled for the first time; 6, the interiors of the holes are metalized for the first time; 7, the through holes are back-drilled for the first time; 8, vacuum hole plugging is carried out; 9, the holes are covered with dry films, single side +5 mil dry film covering is carried out on deep micropores, first-order blind holes and the first through holes; 10, chemical copper reduction and ceramic grinding plate copper reduction are carried out to achieve the size of 20 micrometers to 25 micrometers; 11, second through holes are mechanically drilled; 12, the interiors of the holes are metalized for the second time, and POFV blocking is carried out; 13, the second through holes are back-drilled, and connection type functional holes are machined; 14, an outer layer graph is transferred, and a fine line graph is etched; 15, solder resisting and surface treatment are completed. The high-complexity high-speed PCB can be manufactured, the manufacturing period is short, and the cost is low.

A unified minimum mean square error (MMSE) equalization/multi-user detection (MUD) approach for demodulating direct sequence CDMA (DS-CDMA) signals is described. In at least one embodiment, the unified approach is capable of generating a variety of cost-effective receiver demodulation techniques that may range from, for example, a low cost linear MMSE equalization technique to a relatively high complexity MMSE MUD.