摘要:
Video sequence processing is described with various filtering rules applied to extract dominant features for content based video sequence identification. Active regions are determined in video frames of a video sequence. Video frames are selected in response to temporal statistical characteristics of the determined active regions. A two pass analysis is used to detect a set of initial interest points and interest regions in the selected video frames to reduce the effective area of images that are refined by complex filters that provide accurate region characterizations resistant to image distortion for identification of the video frames in the video sequence. Extracted features and descriptors are robust with respect to image scaling, aspect ratio change, rotation, camera viewpoint change, illumination and contrast change, video compression/decompression artifacts and noise. Compact, representative signatures are generated for video sequences to provide effective query video matching and retrieval in a large video database.
摘要:
Various approaches for motion search refinement in a processing element are discussed. A k/2+L+k/2 register stores an expanded row of an L×L macro block. A k-tap filter horizontally interpolates over the expanded row generating horizontal interpolation results. A transpose storage unit stores the interpolated results generated by the k-tap filter for k/2+L+k/2 entries, wherein rows or columns of data may be read out of the transpose storage unit in pipelined register stages. A k-tap filter vertically interpolates over the pipelined register stages generating vertical interpolation results.
摘要翻译:讨论了处理元件中运动搜索细化的各种方法。 A k / 2 + L + k / 2寄存器存储L×L宏块的扩展行。 k抽头滤波器在扩展的行上水平内插,产生水平插值结果。 转置存储单元存储由k抽头滤波器为k / 2 + L + k / 2个条目生成的内插结果,其中数据行或列可以在流水线寄存器级中从转置存储单元中读出。 k抽头滤波器垂直内插在流水线寄存器级上,产生垂直插值结果。
摘要:
Apparatus and methods for scalable block pixel filtering are described. A block filtering instruction is issued to a processing element (PE) to initiate block pixel filtering hardware by causing at least one command and at least one parameter be sent to a command and control function associated with the PE. A block of pixels is fetched from a PE local memory to be stored in a register file of a hardware assist module. A sub-block of pixels is processed to generate sub-block parameters and the block of pixels is filtered in a horizontal/vertical edge filtering computation pipeline using the sub-block parameters.
摘要:
Various approaches for motion search refinement in a processing element are discussed. A k/2+L+k/2 register stores an expanded row of an L×L macro block. A k-tap filter horizontally interpolates over the expanded row generating horizontal interpolation results. A transpose storage unit stores the interpolated results generated by the k-tap filter for k/2+L+k/2 entries, wherein rows or columns of data may be read out of the transpose storage unit in pipelined register stages. A k-tap filter vertically interpolates over the pipelined register stages generating vertical interpolation results.
摘要翻译:讨论了处理元件中运动搜索细化的各种方法。 A k / 2 + L + k / 2寄存器存储L×L宏块的扩展行。 k抽头滤波器在扩展的行上水平内插,产生水平插值结果。 转置存储单元存储由k抽头滤波器为k / 2 + L + k / 2个条目生成的内插结果,其中数据行或列可以在流水线寄存器级中从转置存储单元中读出。 k抽头滤波器垂直内插在流水线寄存器级上,产生垂直插值结果。
摘要:
Techniques for efficient database formation and search in applications embedded in a media device are provided. The search may be performed synchronously with presentation of media programming content on a nearby media presentation device. A mobile media device captures some temporal fragments of the presented audio/video content on its microphone and camera, and then generates query fingerprints for the captured fragment. A local reference database resides on the mobile media device and a master reference database resides on a remote server with a most recent chunk of reference fingerprints transferred dynamically to the local mobile media device. A chunk of the query fingerprints generated locally on the mobile media device are searched on the local reference database for continuous content search and identification. The method presented automatically switches between the local search on the mobile media device and a remote search on an external search server.
摘要:
Scaleable video sequence processing with various filtering rules is applied to extract dominant features, and generate unique set of signatures based on video content. Video sequence structuring and subsequent video sequence characterization is performed by tracking statistical changes in the content of a succession of video frames and selecting suitable frames for further treatment by region based intra-frame segmentation and contour tracing and description. Compact representative signatures are generated on the video sequence structural level as well as on the selected video frame level, resulting in an efficient video database formation and search.
摘要:
An apparatus and a method for quarter-pel motion compensated search are described in the context of an array processor with tightly coupled, multi-cycle hardware assist attached to each node. A quarter-pel motion compensated search (QPMCS) instruction initiates the quarter-pel motion compensated search pipeline operation. An instruction decode and instruction operation control unit generates a starting address for a 4×4 block of a current macro block search operation indicating where to fetch the pel values. An interpolation unit determines at least eight neighboring quarter-pels per pipeline stage based on the 4×4 block of pel values. An absolute value of difference function computes the absolute value of difference values between a current macro block pel and the at least eight neighboring quarter-pels per pipeline stage. An accumulator accumulates at least eight summation values for the 4×4 block at quarter-pel positions per pipeline stage.
摘要:
A multi-node video signal processor (VSPN) is describes that tightly couples multiple multi-cycle state machines (hardware assist units) to each processor and each memory in each node of an N node scalable array processor. VSPN memory hardware assist instructions are used to initiate multi-cycle state machine functions, to pass parameters to the multi-cycle state machines, to fetch operands from a node's memory, and to control the transfer of results from the multi-cycle state machines.
摘要:
A multi-dimensional database and indexes and operations on the multi-dimensional database are described which include video search applications or other similar sequence or structure searches. Traversal indexes utilize highly discriminative information about images and video sequences or about object shapes. Global and local signatures around keypoints are used for compact and robust retrieval and discriminative information content of images or video sequences of interest. For other objects or structures relevant signature of pattern or structure are used for traversal indexes. Traversal indexes are stored in leaf nodes along with distance measures and occurrence of similar images in the database. During a sequence query, correlation scores are calculated for single frame, for frame sequence, and video clips, or for other objects or structures.
摘要:
Techniques for efficient database formation and search in applications embedded in a media device are provided. The search may be performed synchronously with presentation of media programming content on a nearby media presentation device. A mobile media device captures some temporal fragments of the presented audio/video content on its microphone and camera, and then generates query fingerprints for the captured fragment. A local reference database resides on the mobile media device and a master reference database resides on a remote server with a most recent chunk of reference fingerprints transferred dynamically to the local mobile media device. A chunk of the query fingerprints generated locally on the mobile media device are searched on the local reference database for continuous content search and identification. The method presented automatically switches between the local search on the mobile media device and a remote search on an external search server.