442results about "Image enhancement" patented technology

Method for determining the pose of a camera with respect to at least one object of a real environment for use in authoring/augmented reality application that includes generating a first image by the camera capturing a real object of a real environment, generating first orientation data from at least one orientation sensor associated with the camera or from an algorithm which analyses the first image for finding and determining features which are indicative of an orientation of the camera, allocating a distance of the camera to the real object, generating distance data indicative of the allocated distance, determining the pose of the camera with respect to a coordinate system related to the real object of the real environment using the distance data and the first orientation data. May be performed with reduced processing requirements and/or higher processing speed, in mobile device such as mobile phones having display, camera and orientation sensor.

In some arrangements, product packaging is digitally watermarked over most of its extent to facilitate high-throughput item identification at retail checkouts. Imagery captured by conventional or plenoptic cameras can be processed (e.g., by GPUs) to derive several different perspective-transformed views—further minimizing the need to manually reposition items for identification. One claim recites a system for capturing imagery of retail items supported by a fixture, comprising: at least one camera unit having at least one operative camera to capture imagery of at least a portion of the retail items supported by the fixture; and one or more processors configured to process the captured imagery to discern identifying information associated with the imaged retail items. In one implementation the at least one camera unit comprises a plurality of cameras, wherein a field of view of a first one of the plurality of cameras overlaps with a field of view of a second one of the plurality of camera units. A great variety of other features, claims, combinations and arrangements are also detailed.

Described herein are systems and methods for performing multi-stage detection and classification of cancer regions from digitized images of biopsy slides. Novel methods for processing the digitized images to improve feature extraction and structure identification are disclosed, including but not limited to the use of quaternions, logarithmic mappings of color channels, and application of wavelets to logarithmic color channel mappings. The extracted features are utilized in improved machine learning algorithms that are further optimized to analyze multiple color channels in multiple dimensions. The improved machine learning algorithms include techniques for accelerating the training of the algorithms, making their application to biopsy detection and classification practical for the first time. The performance of the described systems and methods are further improved by the disclosure of a novel multistage machine learning scheme, in which additional classifiers are utilized to choose among the classes proposed by other classifiers in close cases.

The specification and drawings present a new method, system, apparatus and software product for constructing by an electronic device an image panorama from a plurality of images based on their visual quality. The same portion of a panoramic image scene is typically captured on multiple neighborhood video frames from which one can choose the best visual representation to be pasted into the panorama. Constructing the image panorama from the plurality of captured images can comprise the following steps: a) evaluating image quality of K consecutive frames out of the captured plurality of images using a predetermined criterion, b) selecting one image out of the evaluated K frames with the best image quality using a predetermined rule, and c) appending the selected frame to previously selected frames for constructing the image panorama. On-line and off-line implementations are possible.

The invention discloses a hybrid image compressing method based on block classification. In the invention, blocks are divided into text/ image block, natural image block and hybrid block according to different features of text/image area and natural image area of original images in the aspect of color and texture; then a compression mode in which palette coding, Hextile coding and LZW coding are combined is used in the text/ image block, a JPEG lossy compression mode is used in the natural image while the algorithm of the combination of the two modes is used in the hybrid block. The invention has the advantages that based on the block classification method, the text/image area and the natural image area can be divided up accurately, and that the blocks are respectively encoded can keep rather acuminate edge of the text/image block so as to enhance visual effect on the basis of ensuring higher compression efficiency and upgrade the peak signal to noise ratio of reconstructed images; the complexity of the algorithm is lower.

A rotational movement component of a camera provided on a mobile object is to be detected in a precise and simple manner. Three motion vectors extracted from the forward images captured by a camera mounted on a vehicle are extracted on a random basis. A temporary rotation angle is calculated using the extracted motion vectors based on a relational expression that represents the relationship between a background vector and the rotation angle of the camera, the relational expression being a linear expression of the yaw angle, pitch angle, and roll angle of the rotational movement of the camera. An error when using the temporary rotation angle is calculated for other motion vectors, and the number of motion vectors for which the error is within a predetermined threshold value is counted. After repeating such a process for a predetermined number of times, the temporary rotation angle with the largest number of motion vectors for which the error is within the predetermined threshold value is selected as the rotation angle of the camera. The present invention can be applied to an in-vehicle obstacle detecting device.

The present invention provides methods, apparatus, and systems for analyzing a structure using thermal imaging. A plurality of images of a structure are automatically captured using one or more image capture devices. The images may be captured in one or more ranges of wavelengths of light. The images are then processed to generate image data for the images. The image data can then be analyzed to determine one or more properties of the structure. The images may be captured at an angle with respect to the structure of between approximately 45 to 135 degrees. The images may be captured during a time where one of indirect or no sunlight is present.

A system and method are disclosed for presenting projection image information, including a first image generating module or step, for generating a first image representing a first projection of a three-dimensional object; a second image generating module or step, for generating a second image representing a second projection of the three-dimensional object; an image display module or step, for displaying the first and second images; a region selection module or step, for selecting a first region in the first image; a correspondence module or step, for determining a second region in the second image that corresponds to the first region; and a marking module or step, for displaying a first mark on the first image to identify the first region, and a second mark on the second image to identify the corresponding second region.

The invention provides a relevant filtering opposite-thrust target tracking method with an adaptive scale. The method comprises: an initial position and an initial scale of a to-be-tracked target in a video frame are determined, and convolution feature graphs of different layers are extracted respectively by using the initial position as the center and using a deep convolutional neural network; for the extracted convolution feature graph of each layer, tracking is carried out by using a kernel-correlation filtering tracking method to obtain a tracking result; all tracking results are combined by using an adaptive hedging algorithm to obtain a final tracking result as a final position of the to-be-tracked target, so that the to-be-tracked target in the video frame can be localized; after obtaining of the final position of the to-be-tracked target, a final scale of the to-be-tracked target is estimated by using a scale pyramid strategy; and after obtaining of the final position and the final scale of the to-be-tracked target, a to-be-tracked target image block is extracted based on the final scale by using the final position as a center and each kernel-correlation filtering tracking method is trained again to update a coefficient and a template.

A computer-implemented method for tracking a human face in a target video includes obtaining target video data of a human face; and estimating parameters of a target human face model, based on the target video data. A first subset of the parameters represents a geometric shape and a second subset of the parameters represents an expression of the human face. At least one of the estimated parameters is modified in order to obtain new parameters of the target human face model, and output video data are generated based on the new parameters of the target human face model and the target video data.

A registration framework is presented that registers volumetric breast images captured before surgery with intraoperative surface images. The framework may be implemented using either image-based or model-based registration techniques. The method contains the steps of: identifying an air/tissue boundary from a volumetric image created at a first time; processing the volumetric image with an image filter to emphasize the air/tissue boundary; and registering a surface optically scanned image, with the filtered volumetric image, where the surface optically scanned image is created at a second time.

The invention discloses an adaptive histogram reinforced defogging method based on a dark channel. According to the method, a dark channel model is adopted to obtain a dark channel map of a foggy image, the dark channel map is divided into a plurality of square blocks, and a weighting factor of each square block is calculated; then, a sub-block histogram of each square block is subjected to statistics with an implementation method of CLAHE; according to the sub-block histogram, a sub-block limit contrast histogram is calculated; a histogram balanced grey mapping relationship table corresponding to the square block is calculated based on the contrast histogram, and contrast stretching grey mapping is calculated; a final grey mapping relationship table is calculated by using the weighting factor, the grey mapping relationship table and the contrast stretching grey mapping; by using the grey mapping tables of the four adjacent square blocks, a center pixel of each block adopts an original grey mapping relationship, while the other pixels are acquired via grey mapping interpolation of the four blocks, and a defogged image is obtained; and finally, the image is output. According to the adaptive histogram reinforced defogging method based on the dark channel, provided by the invention, the dark channel model and a traditional CLAHE image reinforcement algorithm are combined, so that depth information in a foggy day is effectively utilized and complicated transmission pattern estimation is avoided; a single image is processed; and the adaptive histogram reinforced defogging method is good in defogging effect and high in practicality.

The invention discloses a segmentation method of pathological section unconventional cells based on a multi-scale hybrid segmentation model. The method comprises steps that positive and negative samples are respectively scaled to low-resolution, medium-resolution and high-resolution images, the full convolutional network algorithm is utilized for training to acquire a convergent low-resolution segmentation model, a medium-resolution segmentation model and a high-resolution model; a multi-scale hybrid segmentation model is acquired through fusion by a model integration method; after the effective discriminating area of a new pathological section is processed through utilizing the data enhancement method during testing, the processed effective discriminating area is inputted to the multi-scale hybrid segmentation model, the probability of each pixel in the effective segmentation area is outputted, pixels with probability values greater than the threshold t are taken as abnormal cell pixels and are recorded as 1, remaining pixels are taken as normal cell pixels and are recorded as 0, binary images predicted by the multi-scale hybrid segmentation model are acquired, and post-processingon the binary images is carried out to acquire the final segmentation result. The method is advantaged in that high precision is realized, and a Dice value is above 0.869.

A method to reduce coding artifacts within a discrete decoded picture includes a spatial and/or temporal filtering with respective filter characteristics dependent on an image quality value (Q). Preferably, the spatial filtering includes a deblocking filtering, wherein the deblocking filter operation decreases with an increasing image quality value (Q) and/or the deblocking filtering chooses a image quality value (Q) dependent weighting function for a pixel to be filtered according to an activity of both blocks to which common border said pixel belongs.

Systems and methods for analyzing pathologies utilizing quantitative imaging are presented herein. Advantageously, the systems and methods of the present disclosure utilize a hierarchical analytics framework that identifies and quantify biological properties/analytes from imaging data and then identifies and characterizes one or more pathologies based on the quantified biological properties/analytes. This hierarchical approach of using imaging to examine underlying biology as an intermediary to assessing pathology provides many analytic and processing advantages over systems and methods that are configured to directly determine and characterize pathology from underlying imaging data.

The invention discloses a stacking glass bottle detection method based on machine vision and a transporter based on the method. The detection method comprises the following steps: firstly, obtaining a bottle mouth image sent into a bottle collecting zone via a conveyor belt; then, performing image processing on the obtained bottle mouth graph so as to indentify whether the glass bottle in the current bottle collecting zone has abnormal situation or not; and if yes, giving an alarm. The transporter comprises a glass bottle conveyor belt, a travelling crane provided with a gasbag-type grappling fixture lifting gear and a fastening frame, a camera and a stand column rack, wherein the bottle collecting zone is directly below the fastening frame and is connected with the conveyor belt; a longitudinal guide rail and a horizontal guide rail which can move along the longitudinal guide rail are arranged in the bottle collecting zone; the camera used for collecting glass bottle images in the bottle collecting zone is arranged on the horizontal guide rail; and the lens of the camera is downwards vertical to the bottle collecting zone and is connected with a computer for processing images. The invention is easy to apply, has high degree of automation, high efficiency and wide industrial application prospect.

This invention has as its object to eliminate sensitivity nonuniformity on the plane of an image sensing element irrespective of exchange of a lens. To achieve this object, in an image sensing apparatus which has an image sensing element and can exchange a lens, a memory stores shading correction coefficients respectively corresponding to pixels in a two-dimensional matrix on the plane of the image sensing element, and a correction device makes a correction calculation of pixel data read out from each pixel of the image sensing element by extracting the shading correction coefficient associated with the corresponding pixel from those stored in the memory.

An information processing apparatus includes an image pickup unit, an extracting unit, an estimating unit, a categorizing unit, and a recording control unit. The image pickup unit captures an image of a subject. The extracting unit extracts a human figure from the captured image of the subject captured by the image pickup unit. The estimating unit estimates a posture of the human figure extracted by the extracting unit. The categorizing unit categorizes the posture of the human figure estimated by the estimating unit into a previously prepared pose. The recording control unit controls recording of the captured image on the basis of the pose that the posture of the human figure is categorized into by the categorizing unit.

The invention discloses a method for evaluating the quality of an image shot by an unmanned aerial vehicle, which is characterized in that for an image shot by the unmanned aerial vehicle in patrol inspection, the quality of the image is analyzed from the whole image and local details; for blurring generated by focusing misalignment caused by motions or shakes of the unmanned aerial vehicle, a frequency domain sharpness index and the distribution density and orientation features of edge segments in a space domain are firstly fused so as to judge whether an image captured by the unmanned aerial vehicle image is blur or not on the whole; under the condition that the image quality is acceptable on the whole, salient edge regions in the image are further searched, and the local detail blur degree of image edges is judged through analyzing the average width of difference salient edges. The method is applied to images shot by the unmanned aerial vehicle under outdoor conditions, the image quality can be effectively evaluated under various complex backgrounds without being influenced by the image content, and the evaluation is consistent with subjective evaluation of people in grade.