Abstract: A method of producing an image frame from event packets received from an event camera comprises: forming a tile buffer sized to accumulate event information for a subset of image tiles, the tile buffer having an associated tile table that determines a mapping between each tile of the image frame for which event information is accumulated in the tile buffer and the image frame. For each event packet: an image tile corresponding to the pixel location of the event packet is identified; responsive to the tile buffer storing information for one other event corresponding to the image tile, event information is added to the tile buffer; and responsive to the tile buffer not storing information for another event corresponding to the image tile and responsive to the tile buffer being capable of accumulating event information for at least one more tile, the image tile is added to the tile buffer.

Abstract: A vehicle can include one or more movably mounted cameras that are able to move to adjust a viewing angle of the camera(s) relative to a body of the vehicle. The camera(s) may be movable by virtue of being coupled to a movable side mirror, such that movement of the side mirror changes a field of view of the camera coupled to the side mirror. For example, the side mirror can be rotated about a rotational axis between a first position in which a field of view of the camera is directed in a first direction (e.g. toward a ground proximate the vehicle), and a second position in which the field of view of the camera is directed in a second direction, different than the first direction (e.g., outward from the vehicle or toward a door of the vehicle).

Abstract: An image processing system comprises a template matching engine (TME). The TME reads an image from the memory; and as each pixel of the image is being read, calculates a respective feature value of a plurality of feature maps as a function of the pixel value. A pre-filter is responsive to a current pixel location comprising a node within a limited detector cascade to be applied to a window within the image to: compare a feature value from a selected one of the plurality of feature maps corresponding to the pixel location to a threshold value; and responsive to pixels for all nodes within a limited detector cascade to be applied to the window having been read, determine a score for the window. A classifier, responsive to the pre-filter indicating that a score for a window is below a window threshold, does not apply a longer detector cascade to the window before indicating that the window does not comprise an object to be detected.

Abstract: A method of producing an image frame from event packets received from an event camera comprises: forming a tile buffer sized to accumulate event information for a subset of image tiles, the tile buffer having an associated tile table that determines a mapping between each tile of the image frame for which event information is accumulated in the tile buffer and the image frame. For each event packet: an image tile corresponding to the pixel location of the event packet is identified; responsive to the tile buffer storing information for one other event corresponding to the image tile, event information is added to the tile buffer; and responsive to the tile buffer not storing information for another event corresponding to the image tile and responsive to the tile buffer being capable of accumulating event information for at least one more tile, the image tile is added to the tile buffer.

Abstract: A method of producing an image frame from event packets received from an event camera comprises: forming a tile buffer sized to accumulate event information for a subset of image tiles, the tile buffer having an associated tile table that determines a mapping between each tile of the image frame for which event information is accumulated in the tile buffer and the image frame. For each event packet: an image tile corresponding to the pixel location of the event packet is identified; responsive to the tile buffer storing information for one other event corresponding to the image tile, event information is added to the tile buffer; and responsive to the tile buffer not storing information for another event corresponding to the image tile and responsive to the tile buffer being capable of accumulating event information for at least one more tile, the image tile is added to the tile buffer.

Abstract: A method of producing an image frame from event packets received from an event camera comprises: forming a tile buffer sized to accumulate event information for a subset of image tiles, the tile buffer having an associated tile table that determines a mapping between each tile of the image frame for which event information is accumulated in the tile buffer and the image frame. For each event packet: an image tile corresponding to the pixel location of the event packet is identified; responsive to the tile buffer storing information for one other event corresponding to the image tile, event information is added to the tile buffer; and responsive to the tile buffer not storing information for another event corresponding to the image tile and responsive to the tile buffer being capable of accumulating event information for at least one more tile, the image tile is added to the tile buffer.

Abstract: An image processing apparatus comprises a normalisation module operatively connected across a bus to a memory storing an image in which a region of interest (ROI) has been identified within the image. The ROI is bound by a rectangle having a non-orthogonal orientation within the image. In one embodiment, the normalisation module is arranged to divide the ROI into one or more slices, each slice comprising a plurality of adjacent rectangular tiles. For each slice, the apparatus successively reads ROI information for each tile from the memory including: reading a portion of the image extending across at least a width of the slice line-by-line along an extent of a slice. For each tile, the apparatus downsamples the ROI information to a buffer to within a scale SD<2 of a required scale for a normalised version of the ROI.

Abstract: A convolutional neural network (CNN) for an image processing system comprises an image cache responsive to a request to read a block of N×M pixels extending from a specified location within an input map to provide a block of N×M pixels at an output port. A convolution engine reads blocks of pixels from the output port, combines blocks of pixels with a corresponding set of weights to provide a product, and subjects the product to an activation function to provide an output pixel value. The image cache comprises a plurality of interleaved memories capable of simultaneously providing the N×M pixels at the output port in a single clock cycle. A controller provides a set of weights to the convolution engine before processing an input map, causes the convolution engine to scan across the input map by incrementing a specified location for successive blocks of pixels and generates an output map within the image cache by writing output pixel values to successive locations within the image cache.

Abstract: A vehicle can include one or more movably mounted cameras that are able to move to adjust a viewing angle of the camera(s) relative to a body of the vehicle. The camera(s) may be movable by virtue of being coupled to a movable side mirror, such that movement of the side mirror changes a field of view of the camera coupled to the side mirror. For example, the side mirror can be rotated about a rotational axis between a first position in which a field of view of the camera is directed in a first direction (e.g. toward a ground proximate the vehicle), and a second position in which the field of view of the camera is directed in a second direction, different than the first direction (e.g., outward from the vehicle or toward a door of the vehicle).

Abstract: A dynamically reconfigurable heterogeneous systolic array is configured to process a first image frame, and to generate image processing primitives from the image frame, and to store the primitives and the corresponding image frame in a memory store. A characteristic of the image frame is determined. Based on the characteristic, the array is reconfigured to process a following image frame.

Abstract: An image processing system comprises a template matching engine (TME). The TME reads an image from the memory; and as each pixel of the image is being read, calculates a respective feature value of a plurality of feature maps as a function of the pixel value. A pre-filter is responsive to a current pixel location comprising a node within a limited detector cascade to be applied to a window within the image to: compare a feature value from a selected one of the plurality of feature maps corresponding to the pixel location to a threshold value; and responsive to pixels for all nodes within a limited detector cascade to be applied to the window having been read, determine a score for the window. A classifier, responsive to the pre-filter indicating that a score for a window is below a window threshold, does not apply a longer detector cascade to the window before indicating that the window does not comprise an object to be detected.

Abstract: A vehicle can include one or more movably mounted cameras that are able to move to adjust a viewing angle of the camera(s) relative to a body of the vehicle. The camera(s) may be movable by virtue of being coupled to a movable side mirror, such that movement of the side mirror changes a field of view of the camera coupled to the side mirror. For example, the side mirror can be rotated about a rotational axis between a first position in which a field of view of the camera is directed in a first direction (e.g. toward a ground proximate the vehicle), and a second position in which the field of view of the camera is directed in a second direction, different than the first direction (e.g., outward from the vehicle or toward a door of the vehicle).

Abstract: A method for producing a histogram of oriented gradients (HOG) for at least a portion of an image comprises dividing the image portion into cells, each cell comprising a plurality of image pixels. Then, for each image pixel of a cell, obtaining a horizontal gradient component, gx, and a vertical gradient component, gy, based on differences in pixel values along at least a row of the image and a column of the image respectively including the pixel; and allocating a gradient to one of a plurality of sectors, where n is a sector index, each sector extending through a range of orientation angles and at least some of the sectors being divided from adjacent sectors according to the inequalities: b*16<a*(n+1) where 0?n?5, b=gy and a=gx; and b*16<a*(13?n) where 7?n?12, a=gx+gy and b=gx?gy. At least one sector is associated with a bin; and each instance of a pixel gradient of a cell associated with a bin is accumulated to provide a HOG for the cell.

Abstract: A peripheral processing device comprises a physical interface for connecting the processing device to a host computing device through a communications protocol. A local controller connected to local memory across an internal bus provides input/output access to data stored on the processing device to the host through a file system API. A neural processor comprises at least one network processing engine for processing a layer of a neural network according to a network configuration. A memory at least temporarily stores network configuration information, input image information, intermediate image information and output information produced by each network processing engine. The local controller is arranged to receive network configuration information through a file system API write command, to receive input image information through a file system API write command; and to write output information to the local memory for retrieval by the host through a file system API read command.

Abstract: A multi-processor neural network processing apparatus comprises: a plurality of network processing engines, each for processing one or more layers of a neural network according to a network configuration. A memory at least temporarily stores network configuration information, input image information, intermediate image information and output information for the network processing engines. At least one of the network processing engines is configured, when otherwise idle, to identify configuration information and input image information to be processed by another target network processing engine and to use the configuration information and input image information to replicate the processing of the target network processing engine.

Abstract: A vehicle can include one or more movably mounted cameras that are able to move to adjust a viewing angle of the camera(s) relative to a body of the vehicle. The camera(s) may be movable by virtue of being coupled to a movable side mirror, such that movement of the side mirror changes a field of view of the camera coupled to the side mirror. For example, the side mirror can be rotated about a rotational axis between a first position in which a field of view of the camera is directed in a first direction (e.g. toward a ground proximate the vehicle), and a second position in which the field of view of the camera is directed in a second direction, different than the first direction (e.g., outward from the vehicle or toward a door of the vehicle).

Abstract: A convolutional neural network (CNN) for an image processing system comprises an image cache responsive to a request to read a block of N×M pixels extending from a specified location within an input map to provide a block of N×M pixels at an output port. A convolution engine reads blocks of pixels from the output port, combines blocks of pixels with a corresponding set of weights to provide a product, and subjects the product to an activation function to provide an output pixel value. The image cache comprises a plurality of interleaved memories capable of simultaneously providing the N×M pixels at the output port in a single clock cycle. A controller provides a set of weights to the convolution engine before processing an input map, causes the convolution engine to scan across the input map by incrementing a specified location for successive blocks of pixels and generates an output map within the image cache by writing output pixel values to successive locations within the image cache.

Abstract: An image processing system comprises a template matching engine (TME). The TME reads an image from the memory; and as each pixel of the image is being read, calculates a respective feature value of a plurality of feature maps as a function of the pixel value. A pre-filter is responsive to a current pixel location comprising a node within a limited detector cascade to be applied to a window within the image to: compare a feature value from a selected one of the plurality of feature maps corresponding to the pixel location to a threshold value; and responsive to pixels for all nodes within a limited detector cascade to be applied to the window having been read, determine a score for the window. A classifier, responsive to the pre-filter indicating that a score for a window is below a window threshold, does not apply a longer detector cascade to the window before indicating that the window does not comprise an object to be detected.

Abstract: In one example, a camera is provided that includes: a plurality of MEMS electrostatic comb actuators, each actuator operable to exert a force on at least one lens; and an optical image stabilization (OIS) algorithm module operable to command the plurality of actuators to actuate the at least one lens responsive to motion of the camera.

Abstract: A dynamically reconfigurable heterogeneous systolic array is configured to process a first image frame, and to generate image processing primitives from the image frame, and to store the primitives and the corresponding image frame in a memory store. A characteristic of the image frame is determined. Based on the characteristic, the array is reconfigured to process a following image frame.