Abstract: An electronic device and method for keypoints-based estimation of face bounding box is provided. The electronic device receives an image that includes an object and detects a set of facial keypoints on a face of the object in the image. The set of facial keypoints corresponds to a set of facial features of the object. The electronic device identifies a face pose of the object in the image based on the detected set of facial keypoints and computes a set of values that is associated with a size and a position of a face bounding box in the image. The computation is based on the identified face pose and locations of the set of facial keypoints in the image. The electronic device overlays the face bounding box on the image based on the computed set of values.

Abstract: An electronic device and method for face detection based on facial key-points is provided. The electronic device receives an image of an object of interest. The image may include a face of the object of interest. The electronic device detects a plurality of key-points associated with the face of the object of interest based on the received image and determines a first coordinate value in the received image based on the detected plurality of key-points. Thereafter, the electronic device determines a region in the received image that includes the face of the object of interest based on the determined first coordinate value and controls the display device to overlay a marker onto the determined region of the face. The marker indicates a location of the determined region in the image.

Abstract: To overcome several issues of autofocusing, a deep-learning-based autofocus system utilizes a subject's body to autofocus on the subject's face. The subject, including the subject's body and face/head, are determined utilizing image processing methods, and based on the detection, the subject's face/head is automatically focused on.

Abstract: To overcome several issues of autofocusing, a deep-learning-based autofocus system utilizes a subject's body to autofocus on the subject's face. The subject, including the subject's body and face/head, are determined utilizing image processing methods, and based on the detection, the subject's face/head is automatically focused on.

Abstract: An image processing system, and a method of operation thereof, including: a capture device for obtaining an input image block; and an image signal processing hardware including: a wavelet transform module for generating a transformed image block based on the input image block, the transformed image block having wavelet coefficients, a coding module including a scaling correction module for organizing binary values of the wavelet coefficients into bitplanes and for generating a scaled wavelet block with a scaling factor, the coding module for generating a compressed image block by scanning and coding the bitplanes of the scaled wavelet block, and a decoding module and an inverse wavelet transform module for generating an output image block by decoding, rescaling, and performing an inverse wavelet transform on the compressed image block for display on a display device.

Abstract: Various aspects of a system and method for block-based compression of digital content are disclosed herein. The method comprises quantization, by an encoder, of each block of a plurality of blocks of a digital content based on a computed quantization value. A differential pulse code modulation (DPCM) is computed, by the encoder, to generate quantized residuals of the quantized plurality of blocks. Entropy coding on each of the quantized plurality of blocks is performed by the encoder based on the generated quantized residuals. A compressed bitstream of the digital content is generated by the encoder. The compressed bitstream that comprises depth-first refinement of the entropy-coded quantized plurality of blocks, is generated.

Abstract: An image processing system, and a method of operation thereof, including: a capture unit for obtaining an input image block; and an image signal processing unit coupled to the capture unit, the image signal processing unit including: a pre-processing module for generating a one-dimensional image block from the input image block, a wavelet transform module, coupled to the pre-processing module, for transforming the one-dimensional image block by wavelet coefficients includes a significant partition and a sub-significant partition identified in the wavelet coefficients, a coding module, coupled to the wavelet transform module, for generating a compressed image block including accessing a variable length coding table based on the significant partition and the sub-significant partition, and an inverse wavelet transform module coupled to the coding module for generating an output image block by performing an inverse wavelet transform on the compressed image block for display on a display device.

Abstract: Various aspects of a system and method for image-processing are disclosed herein. The method, implemented in an image-processing device, comprises computation of an average bin value of a histogram for a first color channel of an image. The histogram includes a plurality of bins with a pre-determined bin size, based on pixel values of the first color channel. An adaptive threshold step size is determined for each bin in the histogram, based on a ratio of a bin value of each bin of the plurality of bins and the computed average bin value of the histogram. The method further includes computation of a set of Boolean maps that correspond to a set of threshold values, based on the adaptive threshold step size for each bin. A saliency map is generated based on the computed set of Boolean maps.

Abstract: Various aspects of a system and method for image processing are disclosed herein. The method, implemented in an image-processing device, comprises computation of a plurality of boundary-connectedness values associated with a plurality of regions in a plurality of Boolean maps. The plurality of Boolean maps corresponds to a plurality of color channels of an image. The plurality of boundary-connectedness values associated with the plurality of regions is compared with a pre-specified threshold value. A first set of regions is identified from the plurality of regions in the plurality of Boolean maps as a set of foreground regions, based on the comparison.

Abstract: Various aspects of a system and method for image-processing are disclosed herein. The method, implemented in an image-processing device, comprises computation of an average bin value of a histogram for a first color channel of an image. The histogram includes a plurality of bins with a pre-determined bin size, based on pixel values of the first color channel. An adaptive threshold step size is determined for each bin in the histogram, based on a ratio of a bin value of each bin of the plurality of bins and the computed average bin value of the histogram. The method further includes computation of a set of Boolean maps that correspond to a set of threshold values, based on the adaptive threshold step size for each bin. A saliency map is generated based on the computed set of Boolean maps.

Abstract: Various aspects of a system and method for image processing are disclosed herein. The method, implemented in an image-processing device, comprises computation of a plurality of boundary-connectedness values associated with a plurality of regions in a plurality of Boolean maps. The plurality of Boolean maps corresponds to a plurality of color channels of an image. The plurality of boundary-connectedness values associated with the plurality of regions is compared with a pre-specified threshold value. A first set of regions is identified from the plurality of regions in the plurality of Boolean maps as a set of foreground regions, based on the comparison.

Abstract: An image processing system, and a method of operation of an image processing system thereof, including: a capture device for capturing a first input image frame and a second input image frame; a motion saliency module, coupled to the capture device, for generating a motion saliency frame based on the first input image frame and the second input image frame; a spatial saliency module, coupled to the capture device, for generating a spatial saliency frame based on the second input image frame; and a saliency integration module, coupled to the motion saliency module and the spatial saliency module, for generating the integrated saliency frame based on the combination of the motion saliency frame and the spatial saliency frame, and for determining a region of interest within the integrated saliency frame.

Abstract: An image processing system, and a method of operation thereof, including: a capture device for obtaining an input image block; and an image signal processing hardware coupled to the capture device, the image signal processing hardware including: a coding module including: a scaling correction module for generating a scaled wavelet block based on the input image block, a grey block detection module for determining the input image block to be a grey image block, a compression ratio module for determining the compression ratio of the input image block, a mode setting module for setting a PCM mode identifier based on the compression ratio, the coding module for generating a compressed image block by scanning and coding the input image block using PCM mode; and a decoding module for generating an output image block by decoding the compressed image block for display on a display device.

Abstract: Various aspects of a system and method for block-based compression of digital content are disclosed herein. The method comprises quantization, by an encoder, of each block of a plurality of blocks of a digital content based on a computed quantization value. A differential pulse code modulation (DPCM) is computed, by the encoder, to generate quantized residuals of the quantized plurality of blocks. Entropy coding on each of the quantized plurality of blocks is performed by the encoder based on the generated quantized residuals. A compressed bitstream of the digital content is generated by the encoder. The compressed bitstream that comprises depth-first refinement of the entropy-coded quantized plurality of blocks, is generated.

Abstract: A system and method of operation of an image processing system includes: a pre-processing module for receiving a raw image block of a source image from an imaging device; a transform module for forming a coefficient block by performing a transform operation on the raw image block, and the transform operation including a wavelet transform or a discrete cosine transform; and an encoding module wherein the encoding module is for calculating a coefficient significance vector, a partition significance vector, a run-length coded value, and a Golomb coded value for a portion of a significant partition, and forming an encoded block in an image bitstream having the run-length coded value and the Golomb coded value, and the encoded block.

Abstract: Embedded differential pulse code modulation (DPCM) is performed in response to coding locations of DPCM non-zero residuals during entropy coding utilizing adaptive run-length (ARL) coding. In this manner the need to code zero residuals is eliminated. After adaptive run-length coding, the non-zero residuals are coded in the regular mode, such as utilizing exponential Golomb coding. Overall compression is improved for blocks containing any significant number of zeros.

Abstract: An image processing system, and a method of operation thereof, includes: a pre-processing module for receiving a raw image block of a source image from an imaging device; a wavelet transform module, coupled to the pre-processing module, for forming a wavelet coefficient block by performing a wavelet transform operation on the raw image block; an encoding module, coupled to the wavelet transform module, for generating an encoded block in a wavelet bitplane based on the wavelet coefficient block, for checking a bitstream length of the encoded block in a coding pass, for generating a header in the encoded block based on the bitstream length, and for generating a bitstream based on the header for decoding into a display image to display on a display device.

Abstract: An image processing system, and a method of operation thereof, includes: a pre-processing module for receiving a raw image block of a source image from an imaging device; a wavelet transform module, coupled to the pre-processing module, for forming a wavelet coefficient block by performing a wavelet transform operation on the raw image block; and an encoding module, coupled to the wavelet transform module, for initializing a region significance vector based on the wavelet coefficient block, for generating a code value based on the region significance vector at an index position of a bit region in a wavelet bitplane of the wavelet coefficient block, for forming an encoded block based on the code value, and for generating a bitstream based on the encoded block for decoding into a display image to display on a display device.

Abstract: A system and method of operation of an image processing system includes: a pre-processing module for receiving a raw image block of a source image from an imaging device; a transform module for forming a coefficient block by performing a transform operation on the raw image block, and the transform operation including a wavelet transform or a discrete cosine transform; and an encoding module wherein the encoding module is for calculating a coefficient significance vector, a partition significance vector, a run-length coded value, and a Golomb coded value for a portion of a significant partition, and forming an encoded block in an image bitstream having the run-length coded value and the Golomb coded value, and the encoded block.

Abstract: An image processing system, and a method of operation thereof, including: a capture device for obtaining an input image block; and an image signal processing hardware including: a wavelet transform module for generating a transformed image block based on the input image block, the transformed image block having wavelet coefficients, a coding module including a scaling correction module for organizing binary values of the wavelet coefficients into bitplanes and for generating a scaled wavelet block with a scaling factor, the coding module for generating a compressed image block by scanning and coding the bitplanes of the scaled wavelet block, and a decoding module and an inverse wavelet transform module for generating an output image block by decoding, rescaling, and performing an inverse wavelet transform on the compressed image block for display on a display device.