293 results about "Local illumination" patented technology

Here provides a system, apparatus, and method(s) for processing images based on colour constancy. The method comprises obtaining an image in a first colour space; converting the image to data in a second colour space; transforming the data using chromatic adaptation; performing a first normalisation on the transformed data, where the first normalisation comprises applying a dynamic spatial filtering technique to adjust the transformed data based on light intensity; applying a set of filters to the normalised data, where the set of filters is convoluted based on the normalised data in relation to the image; performing a second normalisation on the filtered data to obtain an illumination estimation of the image in relation to the filtered data; and outputting the normalised data from the second normalisation, where the normalised data maintains colour constancy based on the illumination estimation, removing the illumination from the normalised data.

The invention provides a stack image enhancement processing method, and relates to the technical field of image enhancement, and the method comprises the steps: collecting and preprocessing a stack image, and constructing a training sample set containing a risk level and an inclination angle label; an encoder-classifier dual-stage model is constructed, an encoder extracts features through a multi-scale feature extraction and enhancement module, a space perception attention enhancement module, a dynamic feature pyramid fusion module and a deformation gradient enhancement residual module, and a classifier outputs a risk level and an inclination angle; during training, classification and regression loss weights are dynamically adjusted, an adversarial enhancement sample is generated to enhance robustness, and a gradient is adaptively cut based on a channel attention enhancement vector. According to the method, local illumination interference can be effectively eliminated, multi-scale features are fused, the sensitivity to a sparse risk area is enhanced, and the self-adaptive enhancement processing effect of the heaped image in a complex environment is improved.

An example device includes one or more processors configured to determine a first search area in a first reference picture for a current block of the video data. The one or more processors are configured to determine a first initial reference block in the first search area. The one or more processors are configured to apply a local illumination compensation model to the first search area to generate a refined first search area. The one or more processors are configured to apply template matching to the refined first search area to determine a first candidate motion vector having a lowest template matching cost for the refined first search area. The one or more processors are configured to decode the current block based on the first candidate motion vector.

A video decoder reconstructs a current frame of a video from a video bitstream based on a reconstructed reference frame. For a block of the current frame, the video decoder identifies a reference block in the reference frame based on a motion vector associated with the block. The decoder determines the slope and offset parameters of a local illumination compensation model based on reconstructed pixels in the current frame and the reference frame. The video decoder decodes, from the video bitstream, an adjustment to the slope and updates the slope by applying the decoded adjustment. The decoder further determines an adjusted offset parameter for the local illumination compensation model. The decoder generates predicted pixels for the block by at least applying, to the reference block, the local illumination compensation model with the updated parameters.

The invention discloses a nighttime image enhancement method based on multidirectional information extraction and color loss optimization, which comprises the following steps: firstly, collecting nighttime and normal light road scene images, constructing a nighttime road image enhancement model based on deep learning, and adopting an image enhancement framework based on GAN and a superficial layer semantic enhanced U-Net network as a generator; the five-layer convolutional network is used as a discriminator, then horizontal-vertical bidirectional differential convolution is used to extract multi-directional information of the image, a local illumination attention module is used to balance dark area brightness and bright area overexposure, and finally an adaptive color loss function is used to recover original colors of the image. The method can effectively solve the problems of image quality reduction, detail loss, color distortion and the like under the low-illumination condition, and is particularly suitable for night scene visual perception improvement in the field of vehicle-assisted automatic driving.

Example devices, methods, and computer-readable media for decoding video data are described. An example method includes determining to decode a current block of the video data using a merge mode. The example method includes generating a first merge list for the current block, wherein generating the first merge list comprises applying template matching to candidates of the first merge list. The example method includes generating, based on the first merge list, a second merge list. The example method includes decoding the current block using the merge mode and based on the first merge list or the second merge list.

The invention provides a lightweight target detection method and system based on a multi-modal image, and relates to the field of computer vision, and the method comprises the steps: respectively carrying out the convolution, batch normalization and ReLU activation operation of an infrared image and a visible light image to obtain a feature map, then generating a feature weight mask through the convolution operation and a Sigmoid function, and carrying out the recognition of the feature weight mask; performing pixel-by-pixel weighting on the feature map based on the feature weight mask to generate an optimized feature map; extracting illumination perception information through a local illumination perception module so as to perform weighted calculation on the optimized feature map to obtain a weighted feature map, and splicing the visible light weighted feature map and the infrared weighted feature map in the channel dimension to generate a spliced feature map; and performing subsequent feature extraction operation through the residual connection and the parameter-free channel attention module, and inputting the feature extraction operation to the target detection network for target detection so as to output a detection result. According to the method, target detection with higher precision can be realized on the premise of not introducing larger parameter quantity and calculation quantity.

The invention discloses an intelligent directional irradiation module and an illumination control method.The intelligent directional irradiation module comprises a bottom plate, a plurality of LED light source modules, a microwave detection module, an infrared detection module, a lens moving module and a control module are arranged on the bottom plate, and the distance between the LED light source modules and an irradiated object and the outline of the irradiated object are obtained; the received distance data and contour data are compared with preset target parameters, the lens moving module is controlled to work according to a comparison result, the focal length is adjusted, the size of a light-emitting beam of the LED light source module is changed, an illuminated object is highlighted, white balance on the illuminated surface of the object is reduced, and overexposure is controlled; therefore, the local detail features of the illuminated object are clearly presented, and the illumination effect and the illumination quality of local illumination are improved.

The invention discloses an airport video image classification method based on artificial intelligence, and relates to the technical field of video image classification, and the method comprises the steps: carrying out the whole-course video monitoring of an airport apron region, including a red line region, a stop line and an airplane entry line; carrying out real-time splicing fusion processing on pictures obtained by the video monitoring software by using a video image splicing method; performing frame-by-frame separation on the panoramic video image to obtain each image, namely an airport apron gate panoramic image; a panoramic image of an airport apron parking space is classified, and an illumination perception multi-scale feature pyramid network is used as a core architecture. According to the method, the stability of feature expression is enhanced by fusing local illumination statistics and global channel attention by adopting an adaptive normalization method. According to the method, the deformable convolution kernel group and the scale are adopted to perceive attention on the basis of the feature pyramid network, the receptive field can be adaptively adjusted, the multi-scale features are dynamically fused, and the cross-scale space correlation capturing capability in the image is improved.

Embodiments of the present disclosure provide a solution for video processing. A method for video processing is proposed. The method comprises: determining, for a conversion between a video unit of a video and a bitstream of the video unit, a Local illumination compensation (LIC) type of the video unit based on at least one neighbouring reconstructed sample of the video unit; and performing the conversion based on the determined LIC type.

A device may perform local illumination compensation. The device may determine whether local illumination compensation is to be performed for a luma component of a video block. The device may determine a value of a chroma local illumination compensation variable based on local illumination compensation being performed for a luma component of a video block and one or more thresholds. The chroma local illumination compensation variable may indicate whether local illumination compensation is to be performed for a chroma component of the video block.

Embodiments of the present disclosure provide a solution for video processing. A method for video processing is proposed. The method comprises: obtaining, for a conversion between a current video block of a video and a bitstream of the video, values for a set of adjusting parameters associated with values for a set of model parameters of a local illumination compensation (LIC) model for coding the current video block; updating the values for the set of model parameters based on the values for the set of adjusting parameters; and performing the conversion based on the updated values for the set of model parameters.

A non-transitory machine-readable medium of an electronic device storing computer-executable instructions for decoding a bitstream is provided. The computer-executable instructions, when executed by a processor of the electronic device, cause the electronic device to receive an image frame of the bitstream; determine a first mode from a plurality of local illumination compensation (LIC) modes; determine a first template of a block unit of the image frame and at least one second template of at least one reference block based on the first mode; determine a first model based on the first template and the at least one second template; obtain a slope adjustment parameter; adjust the first model based on the slope adjustment parameter; and modify a prediction illumination of the block unit of the image frame based on the adjusted first model.

A sensing system may include an illumination projector, an image sensor, and one or more controllers. The illumination projector may project a partial illumination pattern on a target object. The image sensor may sense the partial illumination pattern projected on the target object and may generate signals representative of the sensed illumination pattern. The one or more controllers may control the illumination projector and process the signals generated by the image sensor. The illumination projector may include a microlens array and an emitter array. A spatial relationship of emitters of the emitter array to microlenses of the microlens array may cause the emitter array and the microlens array to project light upon the target object in the partial illumination pattern.

Example devices, methods, and computer-readable media are described. An example method includes determining to decode a current block of the video data using a merge mode. The example method includes determining, for the current block, to apply local illumination compensation (LIC). The example method includes determining, for the current block, to apply decoder side motion vector refinement (DMVR). The example method includes decoding the current block based on applying LIC and applying DMVR.

A device may perform local illumination compensation. The device may determine whether local illumination compensation is to be performed for a luma component of a video block. The device determine a value of a chroma local illumination compensation variable based on local illumination compensation being performed for a luma component of a video block. The chroma local illumination compensation variable may indicate whether local illumination compensation is to be performed for a chroma component of the video block.

The invention provides an image enhancement method based on double-domain illumination prior and electronic equipment, and relates to the field of image processing. The method comprises the steps of obtaining a to-be-processed illumination image; according to local illumination distribution characteristics of the to-be-processed illumination image, multi-scale enhancement is carried out on the to-be-processed illumination image after global illumination enhancement, a first illumination image after spatial domain enhancement is obtained, and the local illumination distribution characteristics comprise a brightness mean value and a brightness standard deviation; amplitude information of the to-be-processed illumination image in the frequency domain is enhanced, and a second illumination image enhanced in the frequency domain is obtained; and fusing the to-be-processed illumination image, the first illumination image and the second illumination image to obtain an enhanced target illumination image.

An example device includes memory, and one or more processors configured to determine to decode a current block of the video data using motion information of a second block of the video data. The one or more processors are configured to determine a value of a LIC flag of the second block, the value of the LIC flag of the second block indicative of LIC being applied to the second block. The one or more processors are configured to determine, based on the value of the LIC flag, to apply LIC to the current block. The one or more processors are configured to infer LIC model parameters of the current block to be equal to LIC model parameters of the second block. The one or more processors are configured to decode the current block including using the inferred LIC model parameters and the motion information of the second block.

The invention relates to the field of underground illumination management, and discloses an underground multi-mode illumination collaborative management method and system in a red-mine scene, and the method comprises the steps: capturing the personnel distribution, equipment operation state, environment change and local illumination condition in a working region in real time, employing a time sequence prediction model based on deep learning, and carrying out the real-time monitoring of the operation region, the method comprises the following steps: analyzing and identifying local illumination requirements of different operation areas, predicting an insufficient illumination phenomenon occurring in a short time in the future in combination with a personnel moving track, an equipment moving area and an environmental condition, dynamically adjusting the working state of each illumination unit, calculating a control scheme of each illumination unit by utilizing a collaborative scheduling technology, and performing real-time control. And the output power of the lighting system is dynamically adjusted, an abnormal feedback mechanism is automatically triggered, the dimming strategy is recalculated and adaptive adjustment is carried out, and the effect of the current lighting dimming strategy is analyzed by using a multi-dimensional evaluation method. The method has the advantage of improving the safety of mine operation.

The invention provides an unmanned aerial vehicle remote sensing monitoring and evaluating method and system for field orchard diseases and insect pests, and relates to the technical field of orchard disease and insect pest monitoring and evaluating. The interference of complex background noise is effectively eliminated by constructing a ternary separation index integrating feature enhancement, background suppression and brightness normalization; a threshold value is determined based on the exponential distribution, pixels are accurately divided into a non-canopy background, a shadow canopy and an illumination canopy, and fine segmentation of an unstructured scene is achieved; constructing a gain compensation item by using a local illumination reference mean value, correcting a shadow canopy reflectivity curve, and eliminating illumination non-uniform interference; selecting a local high quantile to establish a health reference basis, coupling a discriminant item and an attenuation item, and resolving to obtain a physiological stress index, so as to realize acute capture of early weak disease characteristics; and performing power weighted accumulation on the abnormal pixel difference value, outputting a regional hazard level index, and objectively quantifying the overall disaster risk of the region through nonlinear aggregation.

The invention relates to the technical field of new energy power system control, in particular to an adaptive power control system based on distributed photovoltaic group dispatching and group control. The self-adaptive power control system based on distributed photovoltaic group regulation and group control comprises a distributed data acquisition module which is used for acquiring illumination intensity, environment temperature, inverter output power and bus voltage of all photovoltaic nodes in real time; according to the method, the threshold value is automatically adjusted along with the environment change through dynamic threshold value calculation and real-time fusion of illumination, temperature and historical voltage deviation data, so that the dynamic threshold value can be synchronously reduced in a photovoltaic shed tree shielding scene when local illumination is suddenly reduced, false triggering caused by threshold value stiffness under a fixed threshold value strategy is avoided, and the reliability of the photovoltaic shed tree shielding scene is improved. And meanwhile, by setting a multi-level power adjustment strategy, a flexible adjustment means is called preferentially, and the photovoltaic unit is cut off only in an extreme scene, so that the loss of power generation is reduced.

A method and apparatus for video coding are disclosed. According to the method, the current block is partitioned into a first region and a second region along a partition line. A target candidate list comprising one or more GPM (Geometry Partition Mode) subblock candidates corresponding to one or more GPM subblock modes for one of the first region and the second region is derived. Said one of the first region and the second region are encoded or decoded using the target candidate list comprising said one or more GPM subblock candidates. According to another method, an inherited LIC (Local Illumination Compensation) flag is determined by inheriting a reference LIC flag associated with a reference block of the current block.

The invention relates to the field of image processing, in particular to a trepanning positioning method for automobile part machining, and the method comprises the steps: obtaining an image of a to-be-trepanned region of a workpiece, and calibrating an ROI region; detecting key points and a main direction of the ROI image by adopting an ORB algorithm; the method comprises the following core steps: calculating the local illumination disturbance degree of each key point neighborhood image block, dynamically adjusting the normalization range according to the local illumination disturbance degree to carry out intensity remapping on the image blocks, and generating ICE-rBRIEF feature descriptors on the basis; matching the generated descriptor with a reference model; establishing a local non-uniform two-dimensional deformation field model by using the high-confidence matching points; and finally, the theoretical trepanning position is accurately compensated, and the trepanning mechanism is guided to work. According to the method, the local illumination disturbance characteristics are analyzed, the intensity normalization range of the image blocks is adjusted in a self-adaptive mode, ICE-rBRIEF feature descriptors are generated, and high-precision trepanning positioning is achieved in combination with deformation compensation.

Describes different implementations, particularly an implementation of video encoding and decoding based on a linear model responsive to neighboring samples. Thus, for a block being encoded or decoded in a picture, refined linear model parameters are determined for a current sub-block in the block, and for encoding the block, local illumination compensation uses a linear model for the current sub-block based on the refined linear model. In a first embodiment, the number N of reconstructed samples increases with the available data for the sub-block. In a second embodiment, partial linear model parameters are determined for the sub-block, and refined linear model parameters are derived from a weighted sum of the partial linear model parameters. In a third embodiment, the sub-blocks are processed independently by LIC.

The invention discloses a modularized multi-light-source shading anti-dazzle structure which is used for a light source module, and a plurality of light sources are arranged on the light-emitting face of the light source module in a matrix mode. The modularized multi-light-source shading anti-dazzle structure comprises a substrate which is installed on the light-emitting face of a light source module through a detachable connecting mechanism, and a plurality of light sources are arranged on the light-emitting face of the light source module in a matrix mode. A plurality of through holes are formed in the substrate, each through hole at least corresponds to one light source in the light source module, and the opening size of each through hole is matched with the overall dimension of the corresponding light source; a light blocking mechanism is arranged on the base plate and comprises a light blocking panel which is fixedly connected to the surface of the base plate and is vertically or obliquely arranged along at least one side of the circumferential edge of the through hole. According to the structure, the light diffusion angle is physically limited through the light blocking panels arranged around the light sources, the illumination area is precisely restrained, stray light overflow is restrained, and the integrated design of high-precision local illumination and low glare is achieved.

Disclosed in the present embodiment are a video coding method and device using bi-predictive local illumination compensation. In the present embodiment, an image decoding device acquires reference pictures of the current picture on the basis of a bitstream. The reference pictures include a reference picture in a direction of LX (X=0 or 1) and a reference picture in a direction of LY (Y=1-X). The image decoding device applies local illumination compensation to the current block regardless of whether the distance between the current picture and the reference picture in the LY direction is the same as the distance between the current picture and the reference picture in the LX direction.

The invention provides an inter-frame prediction method and device, electronic equipment and a computer program product, and relates to the technical field of video coding and decoding. The method comprises the steps that candidate motion information of a current to-be-predicted block is acquired, a reference block corresponding to the current to-be-predicted block is determined based on the candidate motion information, and the candidate motion information comprises an initial identification value corresponding to a local illumination compensation identification; determining a target template of the current block to be predicted, a reference template corresponding to the reference block, and template distortion between the target template and the reference template; performing identification updating operation on the initial identification value according to the template distortion to obtain an updated identification value; and performing inter-frame prediction processing on the current to-be-predicted block according to the update identification value to obtain a predicted value of the current to-be-predicted block. According to the implicit LIC identifier exporting method provided by the invention, on the premise of not additionally increasing the identifier cost, a more accurate LIC opening state is decided, and the prediction error caused by directly inheriting the LIC identifier is reduced.

The invention discloses an image shadow removal method based on multi-scale illumination decomposition, and belongs to the field of image processing, and the method comprises the steps: proposing an improved Retinex-based illumination decomposition network, and decoupling an input image into an illumination component representing illumination distribution and a reflection component containing the inherent texture of an object; a local illumination recovery method based on wavelet features is provided, frequency domain features are extracted by using multi-scale wavelet transform, weight distribution of a shadow area and a non-shadow area is dynamically adjusted through a wavelet attention mechanism, and global illumination consistency is ensured in combination with a reflection component; and a local detail enhancement method based on spatial domain features is provided, the image is further enhanced and refined by using the spatial domain features under the optimized illumination condition, and the texture authenticity of an output shadow removal result is ensured. According to the method and the device, end-to-end shadow removal of the input shadow image can be realized without depending on shadow position prior information, and a high-quality shadow removal result can be obtained.

The invention relates to the technical field of photovoltaic power generation, and discloses a solar panel mounting structure for photovoltaic power generation, which comprises a foundation pile, a base I and a base II, the base I and the base II are connected with each other and both sleeve the top of the foundation pile, and the top of the base I is rotatably connected with a rotating frame I; the other end of the first rotating frame is rotationally connected with a second rotating frame, the top of the second rotating frame is rotationally connected with a mounting frame, the bottom of the mounting frame is rotationally connected with a supporting plate, the other end of the supporting plate is rotationally connected to the top of the second base, the bottom of the mounting frame is further rotationally connected with a sleeve, and the middle of the sleeve is slidably connected with a sleeve rod. Through mutual cooperation of the sleeve, the sleeve rod, the push rod, the first rotating frame, the second rotating frame and other structures, the angle of the mounting frame can be changed so as to better adapt to the local illumination angle, and meanwhile the mounting structure can be folded and stored during transportation, so that the mounting structure is more convenient to carry and mount.

Methods for video decoding and encoding, apparatuses, and non-transitory computer-readable storage media are provided. In one method, a decoder may at least one block vector (BV) based on at least one BV candidate list for an intra template matching prediction (IntraTMP) mode. Additionally, the decoder may obtain a prediction block for a current block based on the at least one BV, or based on the at least one BV and a linear filter of a local illumination compensation (LIC) mode.