13 results about "Image complexity" patented technology

The application relates to a deep learning-based image processing method, device, equipment and medium. The method comprises the following steps: acquiring a to-be-processed dynamic video stream, wherein the to-be-processed dynamic video stream comprises to-be-processed images of each video frame; acquiring a hardware resource state parameter corresponding to a video frame of the to-be-processed image, and calculating a corresponding hardware state value of each to-be-processed image based on the hardware resource state parameter; calculating the image complexity and the interframe difference degree of the to-be-processed image; performing image processing deep learning model state judgment on the to-be-processed image based on the hardware state value, the image complexity and the interframe difference degree, and generating an image processing deep learning model state judgment result of the to-be-processed image. The method can realize hardware adaptation and dynamic model architecture fine-tuning, and significantly improve the model inference efficiency, image processing quality and hardware resource utilization rate of video stream processing.

This invention discloses a method and system for generating digital watermarked images. The method includes: acquiring an original image to be watermarked; generating an image coding mask based on the image complexity of the original image; performing frequency domain decomposition on the original image to obtain a low-frequency component image; embedding an initial watermark image into the low-frequency component image according to the embedding strength values ​​corresponding to the masks in the image coding mask to obtain a low-frequency watermark image; different masks in the image coding mask correspond to different embedding strength values; and performing an inverse frequency domain transform on the low-frequency watermark image to obtain a synthesized watermark image. This invention can improve the success rate and accuracy of watermark extraction.

This application provides a printing timing control method, a printer, a program product, and a storage medium, relating to the field of electronic digital data processing. Utilizing a double-buffered structure of a character buffer and a transmit character buffer, the generation process of print data is physically isolated from the hardware transmission process. This avoids the serial dependency of waiting to transmit and transmitting while waiting to generate, as found in related technologies. It allows the main control module to pre-generate subsequent data and store it in the first-level cache during the hardware transmission intervals. This smooths out fluctuations in data generation time caused by differences in image complexity and avoids hardware stalls caused by computational delays.

This invention proposes a method and system for Thangka image restoration, relating to the field of image restoration. The method includes: preprocessing the acquired original Thangka images; classifying the preprocessed images into first, second, third, and fourth Thangka images according to their complexity from low to high; sequentially performing damage processing on these Thangka images to form structurally damaged, textured, semantically abnormal, and mixed-damaged images; training these damaged images to form a Thangka restoration model; and using the Thangka restoration model to restore the Thangka image to be restored, forming the target restored image. This invention combines the classification and damage processing of different Thangka image complexities to obtain different types of damaged images, adapting to different damage scenarios. Furthermore, by training and optimizing different damaged images separately, the resulting Thangka restoration model can comprehensively restore Thangka images, ensuring that the restored image has a regular structure, natural texture, reasonable semantics, and high fidelity.

This application discloses a video encoding bitrate control method, apparatus, electronic device, and storage medium. The method includes: acquiring the current image frame of a video stream to be processed according to a preset time period; determining each first target region and the image complexity of each first target region of the current image frame according to a preset algorithm; inputting each first target region, the image complexity of each first target region, the target quantization parameter set of the previous image frame, the bitrate of the previous image frame, and the perceived image quality evaluation value of the previous image frame into a parameter fitting model to obtain a preset number of quantization parameter sets for the current image frames; determining the target quantization parameter set of the current image frame based on the preset number of quantization parameter sets of the current image frames and the perceived quality evaluation model; and encoding the current image frame and image frames in the next time period according to the target quantization parameter set of the current image frame.

This invention discloses a visual detection method for defect screening of dried blueberries, specifically relating to the field of visual inspection technology for agricultural products. The method involves acquiring real-time images of the dried blueberries to be tested and segmenting them into individual images; inputting the individual images into a lightweight neural network model to extract basic feature maps; and calculating an image complexity index that fuses content complexity and context complexity in real time based on the basic feature maps; comparing the image complexity index with a preset threshold; dynamically selecting different inference paths and outputting corresponding inference results; this invention dynamically allocates computational resources through the image complexity index, enabling rapid classification of simple images and fine segmentation of complex images, significantly reducing average computation time while ensuring detection accuracy, thus solving the problem of high-precision models being unable to detect in real time; achieving accurate quantification of detection difficulty through the fusion of content complexity and context complexity; and employing a lightweight model design to reduce hardware costs, meeting the speed requirements of industrial sorting.

This invention provides an intelligent watermark replacement method and system based on multimodal feature matching. The method includes: reading the image to be processed; extracting the text, visual, and positional features of the watermark using a multimodal feature parsing module, and outputting a multimodal feature vector; dynamically calculating the OCR confidence threshold and the number of generator iterations based on image complexity and processing feedback to achieve adaptive parameter adjustment; performing watermark generation and background restoration tasks in an integrated manner using an improved attention-gated U-Net generator; transferring the visual style of the original image to the new watermark area using a dynamic style transfer module to achieve automatic visual adaptation; and calculating a visual consistency score through a quantitative evaluation module and feeding it back to the dynamic parameter decision module to form a closed-loop optimization. This invention achieves accurate identification and adaptive replacement of watermarks from complex background images, while ensuring a high degree of consistency between the replacement area and the visual style of the original image.

This invention provides a method for calculating intra-frame code control of I-frames in real-time encoding, comprising: S1, dividing the entire frame image into multiple computational units, with one line of LCUs as one computational unit; S2, determining whether all computational units have been traversed; if yes, the process ends; if no, proceed to step S3; S3, calculating the complexity of the current computational unit; overwriting the complexity of the computational unit at the current position of the previous frame in the buffer with the complexity of the current computational unit; calculating the complexity of the entire frame image; S4, determining whether the current frame is an I-frame; if no, returning to step S2; if yes, proceeding to step S5; S5, calculating the quantization parameter QP of the next computational unit; and returning to step S2. This method relies on historical information for calculation, has good real-time performance, and while meeting the target bit count for I-frames, it requires minimal intra-frame QP adjustment, resulting in stable overall frame image quality and good subjective quality.

This invention discloses an adaptive monitoring load baseline calibration method for dealing with the impact of large passenger flows in subways. The method includes real-time acquisition of multi-dimensional data from subway stations, construction of a correlation prediction model between passenger flow, image complexity, and bitstream, calculation of the theoretical expected bitstream baseline of the monitored video at the current moment, calculation of the relative residual rate between the real-time monitored bitstream and the theoretical expected bitstream baseline, and setting a dynamic anomaly judgment threshold. When the relative residual rate is greater than the threshold, it is judged as mixed attack abnormal traffic, triggering a level-two alarm; when the relative residual rate is less than or equal to the threshold, but the real-time monitored bitstream exceeds a preset proportion of the physical link limit, it is judged as pure business congestion, triggering a level-one warning. This invention can effectively distinguish between malicious attack traffic deviating from the baseline and large passenger flow business traffic conforming to the baseline pattern, enabling the monitoring network to understand and adapt to the normal business scenario of "passenger flow impact," ensuring the continuity of critical monitoring services under extreme passenger flow conditions.

This invention relates to the field of morphological measurement technology, specifically to a system and method for measuring key morphological parameters of a cow's face based on multi-view images. The system includes a data synchronization construction module, a region segmentation and discrimination module, a distortion correction construction module, a pose prediction and fusion module, and a structural parameter measurement module. In this invention, multi-view image data is combined with timestamps, vibration frequency values, and sound wave reflection information for parameter alignment, achieving synchronous integration of image information and external physical states. Confidence is dynamically assessed based on image channels and structural region grayscale values ​​to identify abnormal frames. Image spatial accuracy compensation is achieved by combining pixel offset correction and structural boundary mapping. Spatial direction vectors are constructed using the polarization angle values ​​of the cow's eye region and superimposed with historical behavior to predict pose trajectories. Through multi-view measurement, the stability and accuracy of cow face morphological parameter measurement are effectively improved under conditions of image complexity, pose variation, and structural occlusion.

The embodiment of the application discloses a kind of based on multiscale perception's image complexity prediction method, can reasonably construct receptive field, to improve the accuracy of image complexity prediction.The method uses MPICN to predict image complexity score, in MPICN, MFSA module can be based on the downsampling operation of different sampling multiples and multiple separable convolution receptive blocks, different size features are acquired to the input mixed features, to increase receptive field, and the separable convolution receptive block can be based on the separable convolution layer in different directions to realize position distribution feature extraction to sufficiently perceive the position information of input image, to improve the effectiveness of receptive field, whereby the network can reasonably build enough and effective receptive field.

This invention discloses a real-time processing and scheduling method for infrared thermal imaging based on an FPGA-SoC architecture, specifically relating to the field of thermal imaging system technology. The method acquires raw image frames output by an infrared detector, calculates intra-frame features in real time using an image complexity evaluation algorithm, and dynamically determines the set of image processing algorithms and their priorities based on a complexity and scheduling strategy table. Corresponding processing units are invoked from the FPGA logic resource pool, and the algorithms are executed in parallel or pipelined manner to generate intermediate image data. Image data is transmitted to the SoC shared memory via the AXI4-FULL interface, while control information is written via the AXI4-Lite interface. The RISCV soft core monitors the control register, triggers image output, and refreshes the parameters for the next frame. This method possesses high real-time performance, adaptive scheduling, and hardware-software collaborative control capabilities, making it suitable for infrared thermal imaging applications in complex environments.

This invention relates to the field of image processing technology, specifically disclosing a method and system for natural image ROI segmentation based on eye-tracking gaze duration weighting. Addressing the discrepancy between the segmentation results of traditional methods and the actual attention regions, this method extracts effective gaze points and their gaze durations through eye-tracking data preprocessing. A gaze duration-weighted heatmap is constructed using density peak clustering and a gaze duration weighting model, effectively enhancing the feature representation of high-attention regions. By fusing the eye-tracking heatmap with multimodal semantic features such as color, texture, and edge, the semantic consistency of ROI segmentation is significantly improved. Furthermore, by introducing image complexity adaptive adjustment of the segmentation threshold, flexible adaptation to the segmentation needs of images with varying complexity is achieved. Experimental results demonstrate that the performance of this method and system is significantly superior to traditional methods, thus providing solid theoretical and practical support for intelligent ROI segmentation of high-resolution natural images oriented towards visual attention.