48results about How to "Edge accurate" patented technology

An integrated wet shaving and precision trimming implement includes a handle; a razor cartridge attached thereto; and a trimming tool removably located in the handle. The trimming tool is configured for trimming hair at an edge of a hirsute area. A trimming implement for use during a shaving operation is insertable into a chamber defined by a razor handle and includes a shaft; a blade attached to the shaft; a comb attached over the blade; and a cap attached to an end of the shaft. The cap is receivable on the razor handle to retain the trimming implement in the chamber defined by the razor handle. A trimmer integrated with a handle of a shaving implement includes a reciprocatingly mounted blade and a comb mounted over the blade. The blade is capable of operating in a reciprocating motion via any suitable power source.

The invention discloses a method for segmenting a multi-dimensional texture image on the basis of fuzzy C-means FCM clustering and spatial information and mainly solves the problem of poor quality of image segmentation. The realizing process comprises the following steps of: inputting the texture image to be segmented, carrying out two-dimensional discrete wavelet transformation to the image, and calculating the characteristic vector corresponding to each wavelet coefficient; segmenting the coarsest scale of wavelet transformation; calculating spatial coordinate factors corresponding to the coefficients of the coarsest scale, adding the spatial coordinate factors into an objective function of a traditional FCM clustering algorithm and obtaining the segmenting result marker mapping and the marking field of the scale; obtaining the segmenting result marker mapping of the next scale by adopting the multiple dimensional segmenting method determined by an adaptive scale until the obtained segmenting result marker mapping is at the finest scale; and outputting the segmenting result of the finest scale as the final segmenting result. The method has the advantages of accurate segmenting edge and good consistency of segmenting regions and can be used for segmenting texture images, SAR images including texture information, remote sensing images and medical images.

The invention provides an automatic detection method of static contact angle, including that: a digital camera or CCD sensor is combined with an image acquisition card to capture a water drop image on the plane vertical to the plane where material is located after dripping is completed, and a level set based geometric active contour model method is carried out on the image, so as to obtain a water drop edge; least square fitting is carried out on the edge obtained by the level set every certain iterations, so as to obtain circle parameters, and the obtained minimum curve error is namely the final water drop edge. According to the parameters, tangent lines at junction points of solid, liquid and gas are obtained, thus the static contact angle can be obtained. The invention has the advantages that contact angle can be automatically obtained, accuracy is high and antijamming capability is strong.

The invention discloses an automatic high-accuracy droplet calculation method of dynamic and static contact angles. The method of the invention is realized by obtaining the droplet edge of a shot clear drop image with a level set method and judging whether the image is a static contact angle, an advancing angle or a receding angle. For a static contact angle image, if the droplet is large in size, an ellipse fitting method should be employed, while if the droplet size and the contact angle are both small, a circle fitting algorithm is chosen. If to calculate an advancing angle, a circle fitting or ellipse fitting algorithm is chosen. If to calculate a receding angle, a polynomial fitting algorithm is employed. By obtaining the tangent of a curve resulting from fitting at a droplet marginal point corresponding to a triplet, two contact angles respectively on the left and right are then obtained. The average of the two contact angles is a final contact angle. According to the method of the invention, dynamic and static contact angles can be calculated rapidly and automatically, with high rate of accuracy.

The invention discloses a high-resolution remote sensing image segmentation method based on inter-scale mapping. The high-resolution remote sensing image segmentation method has the advantages that aiming at geographical object extraction involved in objet-level change detection, the high-resolution remote sensing image multi-scale segmentation method based on wavelet transform and an improved JSEG (joint systems engineering group) algorithm is provided; aiming at a key problem that a conventional JSEG algorithm affects segmentation accuracy in high-resolution remote sensing image segmentation, a corresponding improvement strategy is adopted to achieve a good effect; wavelet transform is introduced to serve as a multi-scale analysis tool, excessive rough color quantization of the conventional JSEG algorithm is abandoned, and thus, detail information in original images is kept to the utmost; a novel inter-scale segmentation result mapping mechanism is set up, image segmentation of a current scale is realized on the basis of a segmentation result of a previous scale, and correction of the segmentation result of the previous scale is also realized, and thus, accumulation of inter-scale segmentation errors is decreased effectively; finally, an improved multi-scale segmentation strategy and an improved region merging strategy are provided, and over-segmentation and mistaken merging are effectively reduced.

The invention discloses an image segmentation method based on a wavelet domain hidden markov tree model, which substantially solves the defects existing in a method based on the wavelet domain hidden markov tree model that wavelet domain information is not fully used and a background which guides an image to be segmented is not completely used when back fusing. The processes are that an image block which corresponds to image texture awaiting to be segmented is extracted, training data which corresponds to the image block is extracted, an initial value of an EM algorithm parameter is calculated, a model parameter theta jc which corresponds to the training data is obtained, and a likelihood value jc which is needed in final fusion is obtained. Image multi-scale back fusion segmentation uses different context backgrounds on different scales, and a result on scale 0 is used as a final segmentation result. The invention has the advantages of excellent region homogeneity and accurate edge, can be used for the segmentation which synthesizes an aperture radar SAR image, a remote sensing image and a natural texture image.

Provided are an apparatus and method for converting a low-resolution depth image to a depth image having a resolution identical to a resolution of a high-resolution color image. The depth image conversion apparatus may generate a discrete depth image by quantizing a depth value of an up-sampled depth image, estimate a high-resolution discrete depth image by optimizing an objective functions of the discrete depth image based on the high-resolution color image and an up-sampled depth border, and convert the up-sampled depth image to a high-resolution depth image by filtering the up-sampled depth image when a difference between discrete depth values of neighboring pixels in the high-resolution discrete depth image is less than a predetermined threshold value.

Methods, systems, and apparatus, including computer program products, feature receiving an image mask (1020) where the image mask identifies a first opacity value of pixels in a raster image (1010). The first opacity value ranges between a value representing foreground and a value representing background. Each pixel in the image has a color. A first boundary pixel is identified in the image. A second opacity value for the first boundary pixel is determined using a first local color model of opacity. The first local color model of opacity is determined based on the colors of the pixels in the image within a local neighborhood of the first boundary pixel.

The invention discloses a multi-source remote sensing image pixel-by-pixel classification method based on a correlation fusion network, and the method comprises the steps: reading a multi-spectral image from a data set,wherein the multi-spectral image comprises registered PAN and MS image data and a corresponding class sign true image with only a part of regions; marking an edge sample; respectively preprocessing the PAN image and the MS image, and selecting a training set and a test set; constructing a fusion network model of multi-source remote sensing image pixel-by-pixel classification; constructing a network loss function; training the classification model by using the training data set to obtain a trained classification model; and classifying the test data set by using the trained classification model to obtain the category of each pixel point in the test data set. According to the invention, the classification performance is improved, and the improved loss function strategy improves the pixel-by-pixel classification performance of the remote sensing image.

The invention relates to a visual feature degraded image-orientated solid waste object segmentation method and belongs to fields of robot vision, image segmentation and the like. Due to visual feature degrading and the adhesion and occlusion of solid waste objects, a traditional image segmentation algorithm is difficult to obtain high-precision segmentation results. According to the method of the invention, a background model is obtained through deep background modeling; the background model is compared with solid waste point cloud, so that a foreground mask can be extracted; local masks in the foreground mask are extracted, and therefore, an entire image segmentation problem is converted into a plurality of local mask segmentation problems; and as for the local masks, adhered and occluded objects are segmented through fuzzy region extraction, and fuzzy region re-marking is performed to obtain high-precision segmentation results. The method of the invention has high segmentation accuracy and can effectively segment solid waste objects of which the colors are seriously degraded, and enable ideal segmentation effects for the adhered and occluded solid waste objects.

Disclosed is a segmentation method of a cartoon. The segmentation method of the cartoon includes the following steps: dividing areas of an inputted image: dividing areas to the inputted image according to a user inputting the image and based on textural features, and dividing the areas of the inputted image into a primary area and a non initio area; smoothing the inputted image: smoothing the inputted image based on a texture smoothing algorithm, and obtaining a smoothing chart of the inputted image; extending the primary area: extending the primary area based on the smoothing chart so as to obtain a smoothing connection chart; segmenting the smoothing connection chart: segmenting the smoothing connection chart by using a fingerprint image segmentation method, and obtaining texture areas of the inputted image which are accurately segmented. By comprehensive adopting the above textural features and texture smoothing algorithm, the segmentation method of the cartoon not only has capacity of distinguishing textures, but also can obtain accurate edges, and thereby can segment the accurate texture areas from a cartoon picture.

The invention relates to an image instance segmentation method and device, electronic equipment and a storage medium, and relates to the technical field of computer vision. The image instance segmentation method is used for solving the problem of inaccurate instance segmentation in related technologies. The image instance segmentation method of the present disclosure includes the steps: inputtingthe feature map of the target into a trained segmentation network; based on the segmentation network, determining a sub-image of a target in the feature map, wherein the segmentation network is trained in the following manner; carrying out down-sampling on the sample graph; processing the down-sampled sample graph based on the hole convolution kernels with different expansion rates; determining sample graphs of different receptive fields, fusing the sample graphs of the different receptive fields to obtain a fused sample graph, performing up-sampling and instance segmentation on the fused sample graph to determine a first sub-image, and training an initial segmentation network according to the first sub-image and the marked second sub-image; and carrying out position marking of the sub-images in the image, and because pixel information of pixel points in the feature maps of different receptive fields is large, instance segmentation is more accurate.

The invention relates to a process for coating a strip with a fully-closed outer edge and is mainly applied to the technical field of physical vapor deposition (PVD) coating. The process comprises the following steps: performing computerized numerical control (CNC) machining on the coated film, wherein the profile tolerance is controlled to be 0.1, Cpk is controlled to be more than 1.33 through the SPC control means, and the no-return gauge inspection accords with +0.1/-0.15; coating and cleaning the coated film; putting the coated film into a coating tool, and blocking a non-coated part on the coated film, so that the coated part on the coated film is exposed outside; putting the coating tool in a coating cavity of the magnetron sputtering coating machine, wherein the coating surface is arranged on one side at which a target is positioned; setting the coating process parameters according to the standard; and coating to obtain the coating film with the strip of which the outer edge is fully closed. According to the process, the strip with the fully closed outer edge can be coated, the coating accuracy is high, the film forming position is accurate, the edge is clean and is not shaded, and the cost is low.

PURPOSE: A method and apparatus for positioning the hand in place is provided to decrease an installation space by eliminating the necessity of separately attaching a sensor to an arm robot, and to reduce an interval of transfer time and shorten an interval of duct time by directly transferring a processed article to a next work place. CONSTITUTION: A pair of the first sensors(3) are disposed in the right and left direction of the hand(1). The second sensor(4) is disposed in the side part of the hand. The hand is inserted into an upper or lower portion of the processed member(2) from the front and rear directions. The first sensors calculate an angle of inclination of the processed member in the right and left direction based upon the difference of coordinates of the robot that detects the front edge(2a) or rear edge(2b) of the processed member. The hand is slanted in the right and left direction by the angle(theta) of inclination over or under the processed member. The hand is transferred in the right and left direction until the second sensor detects the side edge of the processed member while the hand is slanted in the right and left direction by the angle(theta) of inclination. The hand is transferred in the front and rear direction by the angle(theta) of inclination until the first sensor detects the front or rear edge of the processed member while the hand is slanted in the right and left direction by the angle(theta) of inclination.

The invention discloses a lens edge grinding machine, and belongs to the technical field of grinding or polishing. The lens edge grinding machine comprises a rack (1) and a PLC control unit, wherein alens fixing device (200), a regulating device (100), a grinding device (300) and a conveying device (400) are arranged on the rack (1); a material rack (500) is arranged on the rack (1); and the regulating device (100) comprises a drive mechanism I (101), a regulator (102) and a rotary rod (103). The lens edge grinding device can freely regulate the angle of a lens, so that the grinding device (300) can precisely grind the edge of the lens, and lens grinding quality is improved; and meanwhile, lens feeding and lens getting further can be simultaneously performed, so that the lens processing and working efficiency is improved.

The invention relates to an RGB-D image saliency detection method, which comprises the following steps of: 1, inputting an RGB image and a depth image into a Resnet50 backbone network in which a full connection layer and a pooling layer are rejected, and extracting RGB features and depth features of each layer; 2, inputting the depth features of each layer into a corresponding depth improvement module, and performing fusion operations such as convolution and activation on the obtained depth improvement features and RGB features to form mixed features of each layer; 3, combining the mixed features in pairs from a high layer to a bottom layer, obtaining saliency maps of all layers through feature grouping supervision fusion modules, wherein the saliency map and a truth value map output by the upper layer are added into each feature grouping supervision fusion module for supervision, and finally the saliency map output by the fourth layer serves as a final result saliency map through iterative optimization. According to the method, noise interference of low-quality depth features is effectively suppressed, high-quality full fusion is performed on the features, a complete and accurate saliency map can be generated, and the method has relatively good detection performance.

The invention relates to a semantic SLAM robustness improvement method based on instance segmentation, and the method comprises the steps: firstly carrying out the instance segmentation of a key framethrough an instance segmentation network, and building prior semantic information; calculating a feature point optical flow field to further distinguish the object, identifying a real moving object in the scene, and removing the feature points belonging to the dynamic object; and finally, performing semantic association, and establishing a semantic map without dynamic object interference. Compared with the prior art, the semantic map is established by adopting a method of combining deep learning and optical flow, and the depth map is added on the basis of the color map, so that the system isendowed with the capability of establishing the dense three-dimensional point cloud semantic map. In addition, a Mask-RCNN framework is adopted for real-time semantic segmentation, and object dynamicinformation can be calculated through mutual combination of dynamic feature points estimated by optical flow information and pixel-level semantic information. According to the method, deep learning and optical flow are mutually combined, so that the robustness of the whole system is remarkably improved, and the method can be applied to real-time semantic map construction in a dynamic scene.

The invention provides a remote sensing image change detection method based on an object-level semi-supervised CV model. The method comprises the steps of preprocessing remote sensing images of all time phases; superposing the images, and performing multi-scale segmentation to form homogeneous image objects; calculating a change intensity characteristic of each image object, and carrying out characteristic mapping on the pixels to obtain a change intensity characteristic graph; performing initial clustering on the object change intensity characteristics to obtain a membership matrix; calculating a category marking information entropy of each object by adopting an information entropy measurement method, and then carrying out category initial marking to generate category marking knowledge; taking the change intensity feature map as an input feature, introducing category marking knowledge into a CV model, constructing an energy functional considering the category marking knowledge, and establishing an object-level semi-supervised CV model; and through solving an Euler equation corresponding to the energy functional, constructing an energy constraint, guiding the rapid evolution of thecurve to the target contour, and realizing the automatic change detection of the remote sensing image.

A method for producing a chemical reactor. The chemical reactor comprises one or more effective channels which comprise pillar structures, an input connected to one of the effective channels to allowfluid/gas into the effective channels and an output connected to one of the effective channels to remove at least one component of the liquid/gas. The method comprises obtaining an initial design of the reactor, further introducing into the initial design at least a structured area positioned adjacent to an effective channel of the one or more effective channels located at the edge of the initialdesign, the structured area not being fluidly connected to one of the effective channels, to obtain a further design and the production of the reactor according to the further design.