Abstract: An image processing system and demosaicing method are provided to calculate estimated missing color sensor values in an image using a linear prediction from the raw color sensor value at the current pixel location. The raw image is divided into regions of sensor values, and the linear relations between color planes for each region are determined by a regression method that calculates the degree to which different color planes co-vary within each region The missing color sensor values per region are calculated as a scaled and shifted version of the raw color sensor values using linear regression coefficients determined from the local linear regression process.

Abstract: An image having higher resolution and/or better color is generated from multiple images of a subject taken by a handheld imaging device. The motion of the device, even when a user holds the device steady, gives the images perspectives that differ by an amount ranging from a fraction of a pixel to a few pixels. The differences in perspective provide different information about the subject that can be combined to create a better image. In particular, an image shifted by a non-integer number of pixels relative to another image provides information about portions of the subject that are between the pixels of the other image and usable for increasing resolution. Images shifted relative to each other by an integer number of pixels have aligned pixels that may provide different color information for the same portion of the subject when the aligned pixels correspond sensor elements having different color filters.

Abstract: An image device, such as a digital camera, detects specific repeating patterns of signal variations by processing columnar information from the device's two-dimensional sensor array used to generate images. In one embodiment, the columnar information is derived from calculating row averages for two image frames, with each row average being a computed average of the multiple signal intensities generated from some or all of the sensors within a particular row. After the columnar information is determined for each of the two frames, a difference signal is generated as a sequence of the differences between the row averages for the first frame and the row averages for the second frame. This row averaging and frame differencing removes a large percentage of the signal energy that is not a result of the artifact of interest, such as the flicker generated by illumination having intensity fluctuations at 100 Hz or at 120 Hz.

Abstract: A method and apparatus for interpolating color image information are provided. One or more image data values for a portion of a digital image in a vicinity of a target pixel are received and stored in a local array. A processor determines whether there is an edge in the vicinity of the target pixel based on the data values in the local array. If there is not an edge in the vicinity of the target pixel, then long scale interpolation is performed on the image data values in the local array, in order to result in interpolating color information that is missing from the image. If there is an edge in the vicinity of the target pixel, then short scale interpolation is performed using image data values in a subset of the local array in a closer vicinity of the target pixel. As a result, accurate color rendering of a digital image is achieved, even in the presence of an edge portion that exhibits great contrast between regions of the image.

Abstract: An image processing system and demosaicing method are provided to calculate estimated missing color sensor values in an image using a linear prediction from the raw color sensor value at the current pixel location. The raw image is divided into regions of sensor values, and the linear relations between color planes for each region are determined by a regression method that calculates the degree to which different color planes co-vary within each region The missing color sensor values per region are calculated as a scaled and shifted version of the raw color sensor values using linear regression coefficients determined from the local linear regression process.

Abstract: A method and apparatus for interpolating color image information are provided. One or more image data values for a portion of a digital image in a vicinity of a target pixel are received and stored in a local array. A processor determines whether there is an edge in the vicinity of the target pixel based on the data values in the local array. If there is not an edge in the vicinity of the target pixel, then long scale interpolation is performed on the image data values in the local array, in order to result in interpolating color information that is missing from the image. If there is an edge in the vicinity of the target pixel, then short scale interpolation is performed using image data values in a subset of the local array in a closer vicinity of the target pixel. As a result, accurate color rendering of a digital image is achieved, even in the presence of an edge portion that exhibits great contrast between regions of the image.

Abstract: An image having higher resolution and/or better color is generated from multiple images of a subject taken by a handheld imaging device. The motion of the device, even when a user holds the device steady, gives the images perspectives that differ by an amount ranging from a fraction of a pixel to a few pixels. The differences in perspective provide different information about the subject that can be combined to create a better image. In particular, an image shifted by a non-integer number of pixels relative to another image provides information about portions of the subject that are between the pixels of the other image and usable for increasing resolution. Images shifted relative to each other by an integer number of pixels have aligned pixels that may provide different color information for the same portion of the subject when the aligned pixels correspond sensor elements having different color filters.

Abstract: A method for interpolating a first color value associated with a first color and a second color value associated with a second color for use in generating a pixel that represents a portion of a digital image, based on a third color value that is associated with a third color. A first matrix is created and stored, comprising first coefficient values that are associated with a other values of the first color for pixels that surround the current pixel of interest. A second matrix is stored that holds second coefficient values that are associated with a second color value. Each of the first coefficient values is bitwise shifted by a pre-determined power of 2. The first color value is created and stored by applying the first coefficient values to the other values of the first color using a bitwise shift operation. The first color value is bitwise shifted by a complement of the pre-determined power of 2.

Abstract: A circuit architecture and method are provided for interpolating a first color value associated with a first color and a second color value associated with a second color for use in generating a pixel that represents a portion of a digital image, based on a third color value that is associated with a third color. Pixel data generated by a digital image sensor is serially received at a register array organized in rows and columns that correspond to pixels of interest that are used in a bicubic interpolation process. Values stored in registers of the register array are coupled to and continuously available to four (4) dot product modules and an interpolator. As the serial data arrives, it is clocked stepwise through the registers, and concurrently used by the dot product modules and interpolator to compute the first color value and the second color value. Data that reaches the end of a line of registers is moved into a corresponding shift register for temporary storage until it is needed again.

Abstract: A method whereby motion can be detected in real time during the acquisition of MRI data. This enables the implementation of several algorithms to reduce or eliminate this motion from an image as it is being acquired. The method is an extension of the acceptance/rejection method algorithm called the diminishing variance algorithm (DVA). With this method, a complete set of preliminary data is acquired along with information about the relative motion position of each frame of data. After all the preliminary data is acquired, the position information is used to determine which lines are most corrupted by motion. Frames of data are then reacquired, starting with the most corrupted frame. The position information is continually updated in an iterative process, therefore each subsequent reacquisition is always done on the worst frame of data. The algorithm has been implemented on several different types of sequences, and preliminary in vivo studies indicate that motion artifacts are dramatically reduced.