72results about How to "Avoid smooth" patented technology

The present invention provides an image processing method. In one embodiment, the method provides discontinuity-preserving image smoothing and segmentation, noise reduction, reduction of image variations particularly those variations caused by illumination conditions, exposure compensation, and dynamic range compression. The present invention also provides an electronic image sensor that is able to detect high dynamic range optical images and produce electronic images with reduced dynamic range.

A reciprocating electric razor head includes a fixed blade and a fixed blade support for supporting and fixing the fixed blade. The fixed blade is provided with hair entrance holes for body hair to enter, and an inner side of the fixed blade is provided with a movable blade capable of moving backward and forward in a reciprocating manner relative to the fixed blade to cut off the body hair inside the hair entrance holes, such that the fixed blade obtains a very flat cutting working face contacting the movable blade without the need for additional precision machining.

The invention discloses a generative-adversarial-network-based blurred face reconstruction method and system. The face reconstruction method comprises: acquiring a training sample and to-be-re-reconstructed data; carrying out generative adversarial network training on two convolutional neural networks by the training sample to obtain an optimized-target-based generative network model; and acquiring a blurred face from the to-be-re-reconstructed data, inputting the blurred face into the generative network model to carry out de-blurring reconstruction of the face, thereby obtaining a clear face.The face reconstruction method and the system have the following beneficial effects: a phenomenon that the image obtained by using an interpolation method is too smooth is avoided; the image recovered by the model becomes clear and more high-frequency details are provided; because no frequency domain method is needed, a phenomenon that the frequency-domain data are lack of correlation is avoided;and the blurred face or shielded face can be reconstructed rapidly and the definition and similarity are high. In addition, the face reconstruction system is composed of an acquisition unit, a modelgeneration unit, and a face reconstruction unit; and the face reconstruction system has the same beneficial effects as the face reconstruction method.

The invention discloses a deep learning-based super-resolution image reconstruction method and system. The method comprises the steps of acquiring an image to be reconstructed and training data; inputting the training data into a multilayer convolutional neural network based on a residual structure for learning; reconstructing an optimal model acquired through input learning of the image to be reconstructed to acquire a super-resolution image. By performing deep learning through multilayer convolution based on the residual structure, the acquired optimal model can be high in super-resolution image reconstruction ability; by structuring the optimal model through a deep learning method, excessive smoothing of images acquired through an interpolation method can be avoided, and meanwhile, high-resolution images restored through the optimal model can be clear and sufficient in high-frequency details; frequency domain methods can be saved, and lack of correlation of frequency domain data canbe avoided. The deep learning-based super-resolution image reconstruction system comprises a target acquisition unit, a training unit and an image reconstruction unit and can achieve advantageous effects identical to those of the deep learning-based super-resolution image reconstruction method.

An image decoding device for decoding a hierarchically encoded compressed code obtained by dividing an image into a plurality of tiles and performing discrete wavelet transform on the pixel values of the image tile by tile includes a tile boundary smoothing part that performs smoothing of tile boundary distortion on the image after the decoding by application of a low-pass filter. The tile boundary smoothing part controls the degree of smoothing of the low-pass filter according to the ratio of the decoding quantity to the entire quantity of the compressed code. The decoding quantity is the portion of the compressed code which portion is to be decoded.

An image decoding device for decoding a hierarchically encoded compressed code obtained by dividing an image into a plurality of tiles and performing discrete wavelet transform on the pixel values of the image tile by tile includes a tile boundary smoothing part that performs smoothing of tile boundary distortion on the image after the decoding by application of a low-pass filter. The tile boundary smoothing part controls the degree of smoothing of the low-pass filter according to the ratio of the decoding quantity to the entire quantity of the compressed code. The decoding quantity is the portion of the compressed code which portion is to be decoded.

The invention discloses a loose tobacco filling method for a bulk curing barn, which comprises the following steps: firstly, manufacturing fixed square bars, a loose tobacco air-distributing partition board and metal contact pins, arranging a tobacco filling bracket and a fixed groove in the bulk curing barn, horizontally placing the air-distributing partition board on the tobacco filling bracket in the curing barn, and inserting the fixed square bars into the fixed groove; and loosing bundled tobacco leaves before tobacco filling, stacking the loose tobacco leaves with leaf tips upward and leaf bases downward, filling the fresh tobacco leaves into two fixed square bars tidily, then inserting the contact pins into holes of the fixed square bars to fix the tobacco leaves, and curing the tobacco leaves after the curing barn is fully filled with the tobacco leaves. The method has the following advantages: (1) the tobacco filling amount is large and is increased by 30 percent compared with a rod hanging method and increased by 10 percent compared with a loose tobacco stacking method, and the utilization rate of the curing barn is high; (2) the method saves labor and energy, saves about 70 percent of labor, and reduces coal and electricity consumption by 20 percent; (3) the method fills the tobacco tidily and reserves even gaps for the tobacco leaf bases, so the method is favorable for discharging moisture and prevent the tobacco leaves form smoothening and not stretching, and the method is easy to control the curing operation; (4) the structure is simple, and the material cost is reduced by 60 percent; and (5) the bulk curing barn is convenient to load and unload.

The invention discloses a point cloud enhancement method based on subsection resampling and surface triangularization. The method comprises the specific steps that an input point cloud (imag file=' 2013107425845100004dest-path-image001. TIF' wi='13' he='32' / ) is divided to obtain the set (img file='752571des t-path-image002. TIF' wi='37' he='24' / ) of the subsets of the point cloud (img file='835430dest-path-image001. TIF' wi='13' he='32' / ), each subset (img file='2012107425845100004dest-path-imag003. TIF' wi='20' he='32' / ) is resampled, data noise is filtered out, and new point sets (img file='dest-path-image005. TIF' wi='23' he='32' / ) which are more even in space distribution are obtained; the sets (img file='250417dest-path-image006. TIF' wi='40 he='32 / ) of all the resampled new point sets are combined, a new point cloud (img file='dest-path-mage007. TIF' wi='16' he='32' / ) is obtained, the surface triangularization is carried out on the new point cloud (img file='114468dest-path-mage007. TIF' wi='16' he='32' / ), and a triangular mesh model ((img file='397682dest-path-image008. TIF' wi='15' he='32' / ) is obtained. According to the method, an environmental structure is restored accurately, and meanwhile the original edges and corners of the environment are prevented from being smoothed by mistake; different sampling densities are selected according to different model surface shape change intensity degrees, and model representation is more efficient. The point sets are projected to a two-dimensional plane to be triangulated at the part of the model, and the calculation efficiency ratio is higher than that of triangularization directly carried out in a three-dimensional space.

An image displaying device having multiple photosensing devices have successfully suppressed a leakage current from each photosensing device and improved the S / N ratio. In the image displaying device, pixels and photosensing devices are disposed as pairs in a matrix pattern on a substrate. Each of the pixels and each of the photosensing devices are driven independently. Each photosensing device includes a semiconductor layer that is a photoelectric conversion layer connected to at least a first electrode and a second electrode. The contact surfaces of the first and second electrodes with respect to the semiconductor layer are disposed so that their center axes are separated from each other.

The present invention provides an image processing method. In one embodiment, the method provides discontinuity-preserving image smoothing and segmentation, noise reduction, reduction of image variations particularly those variations caused by illumination conditions, exposure compensation, and dynamic range compression. The present invention also provides an electronic image sensor that is able to detect high dynamic range optical images and produce electronic images with reduced dynamic range.

The invention relates to a cleansing composition suitable for topical application to human skin, more particularly to an oil-based cleansing composition containing a silicone elastomer gelling agent for removal of make-up from the skin.

The invention relates to a low-rank decomposition and space spectrum constraint-based hyperspectral image time domain change feature extraction method. The method mainly comprises the following stepsof: calculating a difference image between two hyperspectral remote sensing images acquired at a same place and at different times; establishing a low-rank matrix decomposition model with a space spectrum constraint by utilizing internal data structural characteristics of the difference image; and solving each component of the model through an alternate iteration manner so as to extract a time domain change feature. According to the method, the low-rank matrix decomposition model with the space spectrum constraint and a solution algorithm of the model are disclosed by sufficiently utilizing internal structures of data, so that the time domain change feature is effectively extracted, multiple forms of noises are removed, real changes are strengthened and false changes caused by noises are suppressed, thereby improving the change detection precision of time domain change hyperspectral remote sensing images.

A reduction gear capable of satisfying, at high level, each of: realizing a high reduction ratio; preventing backlash by securing rotation smoothness; and carrying out compactification and weight saving of entire mechanism is provided. The reduction gear of the invention is a modified crown reduction gear, comprising: a pressing mechanism 16 operated to rotate; a fixed crown gear 2 fixed to an external member 15; a movable crown gear 1, wherein the difference in the number of teeth between the gears is one; and an output axis 3 flexibly attached to the gear 1. The gear 1 engages with the gear 2 at a slant by pressing-force provided from the mechanism 16. In this instance, contact locations of teeth of the gears 1 and 2 are dispersed at two places existing at both sides between which a gradient center line intervenes.

The invention discloses a driverless lane line detection method based on a generative adversarial network. The driverless lane line detection method comprises three parts of a lane line picture generation network, a discrimination network and a detection algorithm, and is characterized by firstly obtaining the training data of the generation network and a discrimination network, then inputting thelow-resolution image into the generation network to generate a super-resolution image; inputting the high-resolution image and the super-resolution image into a judgment network for definition judgment; and capturing data distribution of the super-resolution image and the high-resolution image according to a judgment result, performing the adversarial training based on the data distribution by the generation network and the judgment network until Nash equilibrium is reached to obtain an optimal generation network, and inputting the super-resolution image generated by the optimal generation network into a detection algorithm to perform lane line identification. The method adopts the generative adversarial network to carry out the driverless lane line detection, and can effectively improvethe lane line detection accuracy.

A reduction gear capable of satisfying, at high level, each of: realizing a high reduction ratio; preventing backlash by securing rotation smoothness; and carrying out compactification and weight saving of entire mechanism is provided. The reduction gear of the invention is a modified crown reduction gear, comprising: a pressing mechanism 16 operated to rotate; a fixed crown gear 2 fixed to an external member 15; a movable crown gear 1, wherein the difference in the number of teeth between the gears is one; and an output axis 3 flexibly attached to the gear 1. The gear 1 engages with the gear 2 at a slant by pressing-force provided from the mechanism 16. In this instance, contact locations of teeth of the gears 1 and 2 are dispersed at two places existing at both sides between which a gradient center line intervenes.

The invention relates to a mushroom bacterial strain preparation method comprising the following steps: selecting a branch of 0.5 cm in diameter and 6-8cm in length, boiling, immersing, washing and draining; evenly mixing the branch with wheat bran and water, and disinfecting the mixture; inoculation; inoculating the branch onto a culture bag, thus forming a bacterial strain material bag after spawn running. The culture bag comprises the following materials (wood chip, sweet osmanthus, daun ngokilo, Sedum aizoon L, wheat bran, cane sugar and gypsum) according to the proportion of 28: 10-15: 10: 5:10:10:20-25:1:1, wherein the water content is 54%, and the weight is 2kg / bag. A conventional method is time wasting and labor wasting in sawdust bacterial strain mode, and high in pollution rate; the branch bacterial strain is long in spawn running period, and late in fruiting; the mushroom bacterial strain preparation method can solve said problems, and is low in bacterial strain disease incidence.

The invention discloses a warping tape surface adhesion material and a preparation method thereof as well as a warping tape and a preparation method thereof. The preparation method comprises the following steps: (1) mixing phenolic resin with a solvent to obtain a mixture M1; (2) adding tetraethylortho silicate into the mixture M1 to obtain a mixture M2; (3) adding nano silicon dioxide, carbon fibers, an antioxidant and an anti-ageing agent into the mixture M2 to obtain the warping tape surface adhesion material, wherein relative to 100 parts by weight of phenolic resin, the use amount of the nano silicon dioxide is 20-50 parts, the use amount of the carbon fibers is 10-30 parts, the use amount of the tetraethylortho silicate is 5-15 parts, the use amount of the solvent is 30-80 parts, the use amount of the anti-oxidant is 0.1-5 parts, and the use amount of the anti-ageing agent is 0.1-5 parts. Due to the design, the service life is greatly prolonged, and the production cost is reduced.

The invention discloses humidification equipment for spinning. The equipment comprises a working platform, idler wheel racks are fixedly connected to the two sides of the bottom of the working platform, and idler wheels are movably connected to the bottoms of the idler wheel racks through rotation shafts; a first supporting rack is fixedly connected to the right side of the top of the working platform, a first roller is movably connected to the top of the first supporting rack through a rotation shaft, and textiles are arranged on the bottom of the first roller; a second supporting rack is fixedly connected to the left side of the top of the working platform. According to the humidification equipment for spinning, through arrangement of a humidifier, an air outlet pipe, a control valve, abox body, a rotation rod, a gear, a toothed plate, a connection pipe, a hose, a connection rod, a sliding block, a universal ball head and a spray head, the humidification efficiency of the humidification equipment for spinning can be high, the humidification height can be adjusted, the equipment is extremely convenient to use during humidification of different textiles, the working efficiency ofusers is improved, and the equipment is beneficial to following processing of the textiles and brings economic benefits to an enterprise.

The invention discloses a denoising method for a remote sensing satellite image, and the method comprises the steps: loading an original remote sensing satellite image, and enhancing the image contrast by adopting an adaptive contrast enhancement method to obtain a to-be-denoised image; Dividing the to-be-denoised image into M * N bounded discretization grids to obtain a discretized to-be-denoisedimage; Detecting pulse noise in the discretized image to be denoised by adopting a median absolute difference judgment criterion to obtain a two-dimensional identification matrix F, wherein pixels inthe two-dimensional identification matrix F are in one-to-one correspondence with pixels of the image to be denoised; Setting a search window and a neighborhood window in the discretized image to bedenoised; And enabling the domain window to slide in the search window and denoising the domain window and by combining a similarity calculation method and a Gaussian weighted average calculation method, obtaining a denoised image. According to the method, excellent texture detail maintenance and a good image denoising effect can be realized, and meanwhile, the algorithm complexity is relatively low. Treatment efficiency is high.

An image displaying device having multiple photosensing devices have successfully suppressed a leakage current from each photosensing device and improved the S / N ratio. In the image displaying device, pixels and photosensing devices are disposed as pairs in a matrix pattern on a substrate. Each of the pixels and each of the photosensing devices are driven independently. Each photosensing device includes a semiconductor layer that is a photoelectric conversion layer connected to at least a first electrode and a second electrode. The contact surfaces of the first and second electrodes with respect to the semiconductor layer are disposed so that their center axes are separated from each other.

The invention relates to the technical field of size detection, in particular to a boiler part size detection method based on image processing. The method comprises the following steps: setting a sliding window to traverse an acquired top view image of a boiler grate, denoising the first window, in a window sliding process, taking a midpoint pixel point of a previous row of a target window as a reference value, constructing a reference index of the target window, and performing window-by-window filtering operation by utilizing the reference index of each target window to obtain a filtering result; obtaining a de-noised image of the top view image; edge extraction is carried out on the denoised image, continuous points in the edge are screened out, to-be-connected edge segments are obtained, and the adjacent to-be-connected edge segments are connected by using the mathematical property of a circle to obtain edge information; performing Hough circle detection on the edge information to obtain an optimal detection circle; and obtaining a characteristic index of the optimal detection circle, constructing a standard degree analysis model according to the characteristic index, and judging whether the current fire grate size is qualified or not. According to the embodiment of the invention, the accuracy of size detection can be improved.

Methods and devices for high-throughput manufacturing of a solar cell with a diode is provided.

The invention discloses a gas cooker with conveniently-adjustable wok stands. The gas cooker comprises a gas cooker body, and the front surface of the gas cooker body is fixedly connected with four supporting frames and provided with slide grooves matched with the supporting frames; and the front surfaces of the supporting frames are fixedly connected with the wok stands, and the surfaces of the wok stands are fixedly connected with supporting seats. According to the gas cooker, through cooperative using of the wok stands, the supporting seats, limiting rods, protective cushions, connection blocks, protective covers, threaded pipes, threaded rods, rotary blocks, limiting slide rods and the slide grooves, the problem that wok stands installed on a traditional gas cooker cannot be flexibly adjusted is solved, the gas cooker with the conveniently-adjustable wok stands has the advantage that the wok stands are adjusted flexibly, the gas cooker can be used for woks of different sizes, normal using of the wok is facilitated, food in the wok can be well heated, in addition, the service life of the wok is greatly prolonged, and a lot of convenience is brought for a user.

The invention discloses a bacteriostatic non-woven fabric and a preparation method thereof. The bacteriostatic non-woven fabric is composed of bacteriostatic ES fiber; the bacteriostatic ES fiber comprises a core layer and a surface layer; the volume ratio of the core layer to the surface layer is 1: (0.5-1); the core layer is prepared from polypropylene; and the surface layer is prepared from the following raw materials in percentage by mass of 90-97% of polyethylene, 0.3-3% of zinc oxide, 0.2-1% of silicon dioxide, 1-3% of a dispersing agent, 1-2% of an antioxidant and 0.5-1% of a cross-linking agent. The zinc oxide bacteriostatic agent is dispersed into the fiber surface layer of the PE / PP fiber, and the bacteriostatic non-woven fabric which is high in cost performance, non-toxic and harmless to the human body and capable of being stably produced in mass is manufactured.

The invention discloses a landscape fallen leaf centralized processing device. The landscape fallen leaf centralized processing device comprises a box body, wherein roller frames are fixedly connectedto the two sides of the bottom of the box body, the bottoms of the roller frames are movably connected with rollers through rotating shafts, the left side of the top of the box body communicates witha feeding pipe, and the left side of the inner cavity of the box body is fixedly connected with a motor. According the landscape fallen leaf centralized processing device, the motor, a fan, a filterscreen, a storage box, pulley frames, the pulleys, a partition plate, a fixing rod, a first connecting part, a movable rod, a second connecting part, a pressing rod, a spring, a fixing plate and an L-shaped pressing plate are cooperatively used, so that the problems that a traditional garden fallen leaf processing method is low in efficiency, and garbage, soil and stones in leaves cannot be distinguished are solved; and the landscape fallen leaf centralized processing device has the advantages of being efficient in classification processing, the next step of processing the fallen leaves is facilitated, the multiple fallen leaves can be compressed, a large amount of use space is saved, and convenience is brought to users.

The invention provides a wire harness peeling device. The device comprises a base; a pulling mechanism, a cutting mechanism and a fixing mechanism are sequentially arranged above the base from left toright; the fixing mechanism comprises an air cylinder I and a fixing block arranged at the lower end of the air cylinder I; the cutting mechanism comprises an air cylinder II and a cutting knife arranged at the lower end of the air cylinder II; an auxiliary knife holder corresponding to the cutting knife is arranged on the surface of the base; the pulling mechanism comprises an air cylinder III,a transverse rod arranged on the right side of the air cylinder III and inward semi-arc-shaped tongs arranged on the front side and rear side of the right side wall of the transverse rod; a waste collection tank is arranged in the base; a waste extraction cabinet which can be extracted from the front side wall of the base is arranged in the waste collection tank; and an opening-closing door is arranged at the upper end of the waste collection tank and is located on the surface of the base and below the semi-arc-shaped tongs. With the wire harness peeling device adopted, working efficiency anda production environment are greatly improved, labor cost is reduced, and production quality is improved.

The invention discloses a multi-layer drawer type automatic material preparation bin based on the RFID technology. The multi-layer drawer type automatic material preparation bin comprises a processing box, wherein a material storage cavity is installed on the inner wall of the processing box, supporting plates are installed on the outer walls of the two sides of the processing box, telescopic rods are installed at the bottoms of the supporting plates, three sets of supporting slideways which are arranged at equal intervals are installed on the inner wall of the processing box and are located above the material storage cavity, supporting frames are installed at the tops of the supporting slideways in an embedded mode, a top cover is installed at the top of the processing box through bolts, and a computer device is installed on one side of the processing box and is located above the supporting plates. Through the arrangement of sensors, a worker can collect redundant wires on a first sensing line and a second sensing line to the inner side of a wire collecting box, the worker can arrange the wires conveniently, the situation that the redundant wires on the device tangle in the automobile driving process is avoided, accidents are avoided, and the driving safety is guaranteed.

The invention discloses an over-temperature warning nut which comprises a nut body, a temperature sensing glass ball fixed to the top end of the nut body, a terminal with the bottom end being fixedlyconnected with the top end of the temperature sensing glass ball, a spring cover, a pull rod, a pressure spring and a warning strip. The terminal is in a cylinder shape with the top end open and the bottom end closed, and a through hole is formed in the bottom end of the terminal. An opening in the top end of the terminal is covered with the spring cover, and the spring cover is fixedly connectedwith the temperature sensing glass ball through the mode that the pull rod penetrates through the through hole. The pressure spring is arranged in the terminal, and the two ends of the pressure springare connected with / make contact with the bottom ends of the spring cover and the terminal correspondingly. The two ends of the warning strip are connected with the inner walls of the spring cover andthe terminal correspondingly. According to the over-temperature warning nut, bolt fastening and over-temperature warning effects can be achieved simultaneously, influences on equipment are avoided, potential safety hazards are eliminated, and applicability is high.

A friction material according to an embodiment is a friction material including: copper in an amount of 0.5 wt % or less; an inorganic material having a cleavage property in an amount of 10 wt % to 20 wt %; a first abrasive material having a Mohs hardness of 6.5 or more and less than 7 and a second abrasive material having a Mohs hardness of 7 or more and 8 or less in an amount of 0.2 wt % to 3 wt %.