24 results about "Spin image" patented technology

A region detection device (1) according to the present disclosure includes a line segment detection unit (32) that detects a line segment from a captured image including an image of a target object having a linear band shape, a main line segment extraction unit (33) that extracts a main line segment (SL) that is a line segment having a length equal to or more than a predetermined value; a correction unit (34) that generates, as a corrected image, a rotated image obtained by rotating the captured image such that the main line segment (SL) is perpendicular or parallel to an arrangement direction of pixels in the captured image, and a region detection unit (35) that detects a target object region indicating the image of the target object from the corrected image.

An image forming apparatus includes a rotatable image bearing member about a rotational axis extending in the axial direction, a charging member, an accommodating portion accommodating toner charged to a predetermined polarity; a developing member, a transfer portion, a static eliminating portion, and a collecting member. In the axial direction, when A is a eliminating width with which the eliminating portion discharges the surface of the image bearing member, B is an effective charging width with which the charging member charges the surface, and C is a collection contacting width with which the collecting member contacts the surface, the formula A<B<C is satisfied. In the axial direction, a surface potential of the surface in an area outside the effective charging width, inside the collection contacting width and in contact with the collecting member is a first potential larger than 0V in the same polarity direction as the predetermined polarity.

The present application relates to the field of image processing technology, and provides an image display method and device. After receiving an image rotation request, the first attribute information of the image to be rotated and the second attribute information of the skybox image are obtained, and the target attribute information is determined. According to the target attribute information, the target image and the target skybox image are drawn respectively. Since the target attribute information combines the first attribute information and the second attribute information, the color difference between the rotated image and the skybox is reduced. In this way, when the displayed target image is rotated, the dizziness caused by the color difference during the image rotation process can be reduced, and the user's immersive experience is enhanced. Moreover, after the rotation is completed, the attribute information of the target image and the target skybox image can be restored, ensuring the normal display of the target image and the target skybox image without affecting the user's normal viewing.

The invention relates to the technical field of microlens detection, in particular to an automatic microlens detection method and system, and the method comprises the steps: obtaining a stripe image of a microlens under the illumination of a stripe structure; preprocessing the stripe image to obtain a preprocessed image with a block area as a foreground; determining a rotation angle according to the boundary of the reference block region, and rotating the corrected stripe image to align the stripe with the row direction; segmenting the rotating image, identifying a connected block, solving an external frame to generate a block mask and a complementary set interval mask thereof, and obtaining a block image and an interval image; performing segmentation on the block image, taking a median value according to rows to replace and generate a block reference image, performing differential binarization to obtain a block defect candidate, and performing independent binarization and merging on an abnormal region; overlapping the interval image and the block mask, performing closed operation to suppress line noise, and then performing binarization inversion to obtain an interval defect candidate; and superposing the two candidate output defect mark images. According to the invention, the defect can be automatically identified and marked and output by the microlens under the stripe background.

The application relates to the technical field of image processing, and discloses a reflection symmetry detection method and system of a two-dimensional image, a terminal and a storage medium.The method comprises the following steps: obtaining an original image in a preset data set, and performing pretreatment and non-symmetry region denoising processing to obtain a non-symmetry region denoising feature map; performing rotation processing on the original image to obtain a target rotation image score map; calculating rotation loss between the target rotation image score map and the non-symmetry region denoising feature map, and obtaining a reflection symmetry axis prediction feature map according to the rotation loss; performing binary mask fusion processing and probability Hough transformation processing on the reflection symmetry axis prediction feature map to obtain an image reflection symmetry axis.The application can effectively improve the accuracy of image reflection symmetry axis detection by performing non-symmetry region denoising processing on the preset data set to obtain a non-symmetry region denoising feature map, and calculating rotation loss between the non-symmetry region denoising feature map and a target rotation image score map.

The present invention discloses an image registration method based on multi-source homogeneous structure point features. First, the phase information of the image is used to extract the feature information of the image, which is not affected by the grayscale change and radiation difference between heterogeneous images, so as to obtain feature points with strong robustness and high stability. Then, for the reference image and the image to be registered, the improved Sobel operator is used to calculate the gradient amplitude and angle information, and a direction histogram is constructed in the circular neighborhood of the feature point, which overcomes the influence of SAR speckle noise interference and avoids over-reliance on grayscale characteristics. Then, the gradient direction corresponding to the main peak of the gradient histogram is selected as the main direction of the feature point, and the gradient amplitude and angle and the main direction are used to obtain a feature descriptor with a high degree of consistency in a logarithmic polar coordinate system. Finally, a fast sampling consistency algorithm is used to eliminate mismatched points, thereby reducing the root mean square error of the algorithm and improving the feature matching accuracy, making the registration of multi-source rotated images more accurate and rapid.

The application discloses a kind of based on motion compensation time domain filtering method for image, comprising: using autoregressive moving average model to the obtained gyroscope error is modeled, the change rule of data is used to estimate model parameter, and gyroscope error model is established;Using the gyroscope data of removing error obtained by gyroscope error model, the rotation matrix of camera is calculated, and the rotation image is compensated in reverse using rotation matrix, and the de-rotation image is obtained.Through kalman filtering to image motion vector sequence, image random jitter is obtained, and the de-rotation image is compensated according to random jitter, and the stable image and the subjective motion vector of image are obtained;Using the motion target detection algorithm in difference map based on large map coordinate system to detect target in stable image.The application has high detection sensitivity, and can detect target and its trajectory quickly through a small amount of frame track correlation.

The application provides a kind of weakly supervised target positioning method and device, belong to target positioning field, method includes: S1: import original image dataset, and original image dataset is divided into image training set and image test set;S2: each original image in image training set is rotated at multiple angles, obtain multiple rotated images of each original image;S3: each original image in image training set and each rotated image of each original image are extracted in multiple channels Feature, obtain the feature spectrum group of each original image in image training set and the feature spectrum group of each rotated image;S4: according to the preset layer feature spectrum in the feature spectrum group of each original image in image training set and the preset layer feature spectrum in the feature spectrum group of each rotated image, the calculation of rotation prediction total loss function is carried out, and the rotation prediction total loss function is obtained.The application can improve the accuracy of weakly supervised positioning by mining the complementarity and consistency of fine-grained targets.

The invention relates to a sea surface anti-flare method based on rotation polarization of an unmanned aerial vehicle, and the method comprises the steps: arranging a rotatable polarizing film in front of a lens of an image collection device, driving the polarizing film to rotate through a stepping motor, and enabling the image collection device to collect an image at a fixed frequency along with the rotation of the polarizing film; all the images collected by the polaroid in one rotation period are evaluated, the saturated pixel proportion, the contrast ratio and the information entropy of each image are calculated, a comprehensive evaluation function is established, the comprehensive evaluation functions of all the images in the period are compared, and the image with the optimal comprehensive evaluation function is screened out to serve as the optimal anti-blazing image. According to the method, the motor drives the polaroid to continuously rotate back and forth at a high speed, it is ensured that the polarization angle range of the polaroid is traversed in an extremely short time, image frames with the vibration transmission direction of the polaroid perpendicular to the main polarization direction of flare light can be captured inevitably, artifacts or information loss occurring in a traditional digital image processing method is avoided, and the image quality is improved. And efficient and reliable flare suppression capability is realized.

The invention relates to the technical field of pointer meter reading identification, in particular to a square pointer meter reading identification method and system, and provides the following scheme: obtaining a dial image of a target square pointer meter; performing multi-angle rotation processing on the image to generate a rotation image covering different scale sections; cutting the rotating image to obtain a target image area comprising a scale frame and a number frame; determining an image area to which the pointer belongs and the position of the pointer relative to the scale line through edge detection; compared with the prior art, the scale value indicated by the pointer is calculated based on the pixel distribution, the reading accuracy and the recognition stability of the square pointer meter under the conditions of image skew, pointer criticality and the like are improved, and the method is suitable for various industrial and equipment monitoring scenes.

The invention discloses a space target rotation motion parameter estimation method based on an ISAR sequence image, and the method comprises the steps: obtaining an effective rotation vector of a target relative to a radar according to the ISAR sequence image, carrying out the interpolation calibration of a first image, and obtaining a first image corresponding to the first image; rotating a first image corresponding to the first frame of image to obtain a first rotating image; rotating the first image corresponding to the first frame of image to obtain a second rotating image; obtaining a first similarity according to the first rotation image and a first image corresponding to the first frame image, obtaining a second similarity according to the second rotation image and the first image corresponding to the first frame image, and calculating to obtain a similarity calculation result of the first frame image; calculating a similarity calculation result of each frame of image to obtain a target rotation motion parameter; and processing the target rotation motion parameter by adopting an optimization algorithm to obtain an optimal target rotation motion parameter. According to the method, accurate estimation of the target rotation parameter is realized only depending on the ISAR sequence image.

The present invention discloses a super-resolution imaging method and device, which relate to the field of imaging processing technology. The super-resolution imaging method provided by the present invention uses an adaptive moment estimation method to perform end-to-end training on an initial network model based on a constructed training data set to obtain a trained network model. Then, a multi-rotating camera is used to obtain a rotated image of the target to be imaged, and the rotated image is input into the trained network model to obtain a high-resolution image of the target to be imaged. It can be seen that the present invention adopts a rotating camera method and introduces a deep neural network model of the imaging physics process, which can effectively encode and decode high-frequency information in low-resolution images and output a high-resolution image that restores more high-frequency information, thereby improving the image resolution while improving the imaging quality.

The present invention relates to the field of image segmentation, and provides a method and device for positioning and segmenting the detection area of a casing based on a Bezier curve. The method includes annotating and generating a Bezier curve template, and selecting a detection area according to the Bezier curve template; performing mean filtering and bilateral filtering on the casing image to be detected to obtain image I c ; obtaining the upper and lower boundaries of the casing image according to image I c ; rotating image I c so that the sum of the angles between the upper boundary, the lower boundary of the rotated image I c and the horizontal line is 0, and denoting the rotated image as image I R ; obtaining the weighted edge image of image I R by using a multi-threshold Canny edge detection algorithm; and obtaining the detection area of the casing image to be detected by matching the Bezier curve template with the weighted edge image. The present invention can reduce the workload of annotation, improve the segmentation accuracy, and be applicable to various casing images.

The invention relates to the technical field of image processing, and discloses a two-dimensional image reflection symmetry detection method and system, a terminal and a storage medium, and the method comprises the steps: obtaining an original image in a preset data set, carrying out the preprocessing and asymmetric region denoising processing, and obtaining an asymmetric region denoising feature map; performing rotation processing on the original image to obtain a target rotation image score graph; calculating the rotation loss between the target rotation image score graph and the asymmetric region denoising feature graph, and obtaining a reflection symmetry axis prediction feature graph according to the rotation loss; and carrying out binary mask fusion processing and probability Hough transform processing on the reflection symmetry axis prediction feature map to obtain an image reflection symmetry axis. According to the invention, the asymmetric region denoising processing is carried out on the preset data set to obtain the asymmetric region denoising feature map, and the rotation loss between the asymmetric region denoising feature map and the target rotation image score map is calculated, so that the accuracy of image reflection symmetry axis detection can be effectively improved.

The invention discloses an image rotation method based on angle partition fitting and adaptive adjustment, and the method comprises the following steps: adaptively selecting a corresponding sliding window through an image adaptive recognition adjustment strategy, dynamically extracting feature primitives, forming a feature primitive library, and dynamically updating the feature primitive library through a primitive feature evolution model, storing the elements subjected to high-frequency multiplexing; averagely dividing 360 degrees into N intervals, obtaining fitting parameters of cos theta and sin theta of each interval by adopting an angle coefficient segmentation fitting strategy, and storing the fitting parameters; generating a rotation image of an adjacent primitive according to the coordinate offset of the rotation result of the primitive in the feature primitive library; according to the image rotation method based on angle partition fitting and adaptive adjustment, image region features can be intelligently recognized by introducing an image complexity detection technology, accurate allocation of computing resources is achieved by combining an angle coefficient segmentation fitting strategy, and compared with a traditional scheme, more than 70% of logic resources and compensation coefficients at special angles are saved.

This invention discloses an image rotation device and method. The device includes a hardware rotation unit, a hardware scaling unit, and a software unit. Under the control of the software unit, the hardware rotation unit and the hardware scaling unit perform rotation processing on the original image and output a processed image. The rotation processing includes lossless rotation processing and stretching rotation processing. During lossless rotation processing, the hardware rotation unit rotates the pixels of the original image at the same angle, keeping the relative positions of the pixels unchanged, thus obtaining a lossless rotated image. During stretching rotation processing, the hardware scaling unit stretches or compresses the lossless rotated image, making the size of the stretched or compressed lossless rotated image the same as the size of the original image. The image rotation device and method of this invention improve rotation efficiency, quickly meet rotation requirements, and provide smoother image display, avoiding stuttering.

An image processing apparatus generates a corrected image to correct distortion in a captured image shot by an image capturing apparatus and obtains an internal parameter of the capturing apparatus based on a straight line edge in the corrected image. For parameter candidates serving as correct candidates for the internal parameter, a corrected image is generated by performing a perspective projection conversion on the captured image using the parameter candidate. One of the plurality of parameter candidates is obtained as the internal parameter of the capturing apparatus based on a straight line edge in each corrected image. For each of a plurality of rotated images obtained by rotating the corrected image by different rotation angles, obtained are a length of a vertical straight line edge in the rotated image, an evaluation value becoming higher as a total of the lengths obtained increases, and a parameter candidate providing a highest evaluation.

This application discloses an image correction model training method, image correction method, device, and storage medium. The model training method includes: acquiring feature information of an original image and acquiring a reference image corresponding to the original image; rotating the original image to obtain multiple rotated images, acquiring the inversion information of the rotated images, and constructing training samples from the multiple rotated images and their inversion information, along with the corresponding reference image; training a first preset neural network for predicting the inversion recognition information of the rotated images using the training samples; after training, inputting the rotated images into a second preset neural network to obtain a first corrected image; inputting the rotated images into the converged first preset neural network to obtain the inversion recognition information of the rotated images; obtaining a second corrected image based on the inversion recognition information and the first corrected image; completing the training of the second preset neural network to obtain an image correction model. Therefore, the complexity of the image correction model is reduced by recognizing the inversion information.

The present invention relates to a matching method, device, and storage medium for significantly rotated images. The method comprises the following steps: inputting a reference image and a target image and performing preprocessing; based on the inherent translation invariance of a convolutional neural network, simplifying the rotated image matching problem to a corresponding point matching problem using a group equivariant neural network modeled based on the plane rotation group SO2; constructing and training a rotation equivariant image feature extractor; establishing two matching routes, and applying a nearest neighbor matching algorithm to the image features of the images in each matching route; and performing a primary filtering, backprojection, merging, and secondary filtering on the matching results of the two matching routes using a dual rotation equivariant excitation mixer, to output a matching relationship of dense corresponding points in the reference and target images. Compared with the prior art, the present invention can overcome the effects of significant illumination changes and achieve image matching under all rotation angles.

The present disclosure provides a training method and related device for a region detection model in a spinning process. It relates to the field of artificial intelligence technology. The method comprises: sampling from a spinning task data set and a playback memory to obtain mixed data; inputting the mixed data set into a student model and a pre-trained teacher model respectively to obtain a first prediction result output by the teacher model and a second prediction result output by the student model; for a spinning image sample, the student model is used to detect a candidate fault region in the spinning image sample; cropping the candidate fault region from the spinning image sample to obtain a first sub-image; inputting the first sub-image into a yarn fault detection model to obtain a classification result of the first sub-image, the classification result including a floating yarn discrimination result and a broken yarn discrimination result; optimizing the parameters of the student model based on a first loss determined by the first prediction result and the second prediction result, and a second loss between the classification result and the classification label to obtain a region detection model.

The invention relates to the field of image processing and defect detection, and particularly provides an industrial material defect detection method and device. S1.1, automatically rotating the image, and adjusting the angle of the image according to the outline of the whole target; s1.2, converting the image into a gray-scale image, and then calculating a histogram of the obtained gray-scale image; s1.3, carrying out expansion operation on the image; s1.4, a findContours () method is adopted to comprehensively detect all contours in the image; s1.5, the minimum enclosing rectangle of the corresponding contour is calculated; detecting defects; s2.1, carrying out accurate region division; s2.2, an inRange () function in the OpenCV library is used for creating a binary mask; and S2.3, marking the defect in the original image in a visual mode. Compared with the prior art, the method has the advantages that the accuracy and convenience of defect detection are remarkably improved, and the method has wide application prospects and practical value.

The invention discloses a nuclear industry CT image fine-grained defect identification method based on feature domain rotation consistency and dynamic constraint, and belongs to the technical field of computer vision and image processing. The method comprises the following steps: firstly, carrying out feature extraction on an original image and a rotating image thereof by utilizing a Vision Transform network to obtain corresponding Patch-level feature representation; and then restoring the Patch feature sequence into a two-dimensional feature map, and executing a rotation alignment operation in a feature space. Then, channel compression is carried out on the two-dimensional feature map after rotation alignment, local Patch features with overlapped areas are extracted in a sliding window mode, and Patch feature pairs corresponding to the spatial positions one by one are constructed; further, a dynamic threshold prediction mechanism is introduced, consistency constraint strength is adaptively set for different local areas, and only penalty is applied to feature differences exceeding a threshold, so that a rotation consistency loss function based on the dynamic threshold is constructed.

The embodiment of the invention provides a drug use detection processing method and device.The drug use detection processing method comprises the steps that in the drug use detection process, after a drug image obtained by conducting image collection on a drug before drug use is obtained, drug use detection is conducted in combination with the drug image and at least one pre-configured drug use task, after medicine use detection is passed, medicine use state information, recognized based on the rotation image sequence of the medicine after medicine use, of the user for the medicine is obtained, medicine use feedback processing is conducted on the associated user in combination with the medicine use state information and the corresponding medicine use feedback level, and therefore medicine use state recognition is achieved through the rotation image sequence of the medicine after medicine use; therefore, medicine use feedback is carried out to the associated user through the medicine use state information obtained through identification.

The invention relates to the technical field of image processing, discloses a medical instrument image processing method, and particularly relates to an endoscope rotation image processing method and an endoscope. According to the invention, the spatial position of the real-time image is updated to obtain the rotation image, the problem of image rotation caused by rotation of the endoscope is solved, and stable display of the image is realized. The parameters of the feature extraction module are dynamically adjusted by determining the rotation angle distribution result, so that the feature extraction process can adapt to the rotation change degree of the endoscope, and the accuracy and comprehensiveness of feature expression are improved. Moreover, image estimation is iteratively updated based on a regularization item of a convolution attention mechanism, spatial frequency information is captured in a convolution mode, a key area is focused in combination with the space and channel attention mechanism, image blurring is effectively removed, and the image restoration quality is improved. In addition, a pure rotation angle is accurately separated from the compound motion by an attitude calculation method based on quaternion and a spiral displacement method, so that the precision of rotation correction is ensured.