558 results about "Color model" patented technology

A system and method are provided for tracking a face moving through multiple frames of a video sequence. A predicted position of a face in a video frame is obtained. Similarity matching for both a color model and an edge model are performed to derive correlation values for each about the predicted position. The correlation values are then combined to determine a best position and scale match to track a face in the video.

A method, system, and computer-readable storage medium for automatic segmentation of a video sequence. A segmentation shape prediction and a segmentation color model are determined for a current image of a video sequence based on existing segmentation information for at least one previous image of the video sequence. A segmentation of the current image is automatically generated based on a weighted combination of the segmentation shape prediction and the segmentation color model. The segmentation of the current image is stored in a memory medium.

Methods, apparatuses and systems for tooth shade analysis and matching are disclosed. A method embodiment according to the invention includes acquiring at least one image, the image including one or more teeth of a patient and normalization references, normalizing the at least one image in accordance with the normalization references, determining the color of the teeth as composed of colors from one or more selected shade standards, communicating the standardized color information to a dental laboratory, manufacturing a dental prosthesis based on the standardized color information and installing the dental prosthetic. Normalization is performed to correct patient images for variations in lighting conditions and image source. Normalization references include a black reference, a white reference and at least one color reference. The normalized image is standardized by matching the pixels of the normalized image to selected shade standards. Selected shade standards may include incisal tooth shade standards. The standardized color image and / or analysis may be communicated to a dental laboratory according to aspects of the invention. The dental prosthesis can then be manufactured by a lab technician by referring to the standardized image. Color models according to the invention may include RGB, HSI and other models for representing color images. The methods, apparatuses and systems for tooth shade analysis and matching are also applicable to direct restorations of natural teeth, such as repair of chipped or broken teeth.

A color selection and coordination system including a database of predetermined color relationships implementing a data-driven color model. A starting color is associated with a first color in the color database. One or more predefined color palettes associated with the first color are retrieved. Each palette including one or more coordinating colors, the coordinating colors being predetermined based on the first color and a color coordination algorithm.

A method, system, and computer-readable storage medium for applying Gaussian Mixture Models (GMMs) to local image patches using an adaptive color lookup table. Per-channel color quantization may be performed to find representative colors for a local image patch. Each combination of the representative values corresponds to a representative color. The probabilities of the representative colors may be computed using a local GMM color model and stored to corresponding entries in an adaptive color lookup table. For every pixel in an image patch, the closest representative color may be found, and the corresponding probability may be retrieved from the lookup table and used for the pixel. The method may, for example, be applied to each local window in a method for automatically determining segmentation in a digital video image sequence to calculate the foreground probabilities for the pixels in a propagated classifier via a GMM.

The present invention is embodied in a system and process for automatically learning a reliable color-based tracking system. The tracking system is learned by using information produced by an initial object model in combination with an initial tracking function to probabilistically determine the configuration of one or more target objects in a temporal sequence of images, and a data acquisition function for gathering observations relating to color in each image. The observations gathered by the data acquisition function include information that is relevant to parameters desired for a final color-based object model. A learning function then uses probabilistic methods to determine conditional probabilistic relationships between the observations and probabilistic target configuration information to learn a color-based object model automatically tailored to specific target objects. The learned object model is then used in combination with the final tracking function to probabilistically locate and track specific target objects in one or more sequential images.

A flame detecting method is provided. The method comprises steps of capturing a video segment for a object; analyzing if an image of the object is moving; analyzing at least one of a color model and a flickering frequency of the moving object; comparing analyzed results obtained from the analyzing steps to a flame feature; and determining if the object is a flame.

The invention relates to a method for fire identification through infrared and visible-light video image fusion. A fire is identified through an infrared image preprocessing module, a suspicious infrared image judging module, an infrared image flame determining module, and a visible light fusion comprehensive judgment module. The infrared image preprocessing module is used for carrying out graying treatment, region connection, binarization and dilation operation on an image. The suspicious infrared image judging module is used for calculating the difference between the gray-scale map and the binary image to determine a suspicious image of which the scene changes. The infrared image flame determining module is used for calculating the similarity between two sequential frames of suspicious image to determine whether there is flame. The visible light fusion comprehensive judgment module is used for judging whether the RGB value of a corresponding visible-light image meets a color model of flame based on the model of the flame in an RGB color space and finally judging whether there is a fire. Fusion identification of infrared and visible-light images is realized, missing and false fire reporting is avoided, and the accuracy of fire identification is improved.

The invention relates to a winter jujube disease identification method based on a deep convolutional neural network and a disease image. The method includes the following steps that: an original winter jujube diseased fruit RGB color image collected by an Internet of Things is converted into a YUV color model, and preprocessing is performed; the a rectangular region of interest of disease spots which contains a disease image is extracted, the rectangular region of interest is segmented through using a K-means clustering algorithm, so that a YUV color disease spot image can be obtained; and a three-channel layered convolutional neural network model is constructed and is trained by using training data, and a jujube disease image to be recognized is inputted into the trained model so as to be subjected to disease category identification. The method of the invention can be applied to an Internet of Things-based greenhouse winter jujube disease monitoring system and can obtain accurate disease identification results.

The invention relates to a method for monitoring a moving object in a natural environment. Based on space-time information fusion and feature identification, the method mainly comprises background analysis, foreground extraction, shadow removal and background update. According to the thought of the method, image sequences are converted into hue, saturation and value (HSV) color models, the change rule of the parameter of each pixel point in a certain period of time is analyzed and compared, and each pixel point belonging to a background point, a moving object point or a shadow point in a certain frame can be distinguished through judgment of a judgment formula. By adopting the method, an algorithm is improved on the influence possibly caused by natural conditions such as wind, sunlight and lightning, and the background can be intelligently updated under the conditions that the illumination is abruptly changed, the moving object is still and then is blended into the background, the background object moves and the like; and the method is applied to monitoring of the moving object in the natural environment.