41 results about "Background factors" patented technology

The invention discloses a visual attention and mean shift-based target detection and tracking method. The method comprises the following the steps: firstly, extracting the salient region of a first frame image in an image sequence by using a visual attention method, and removing interferences of background factors to obtain a moving target; then, changing fixed bandwidth of a kernel function in a traditional mean shift method into dynamically changed bandwidth, and tracking the detected moving target by using an improved mean shift method. Shown by an experimental result, the visual attention and mean shift-based target detection and tracking method disclosed by the invention is suitable for infrared and visible image sequences, and better in tracking effect. Moreover, the positional information of the moving target also can be provided by the visual attention and mean shift-based target detection and tracking method disclosed by the invention, and thus, possibility is provided for accurate positioning of the target. The visual attention and mean shift-based target detection and tracking method has a broad application prospect in the military and civil field of night vision investigation, security and protection monitoring, and the like.

The invention discloses a method for measuring influence of background factors to infrared temperature measurement. The method comprises the following steps of: memorizing calibrated experimental data into a software system for measuring the influence of the background factors to the infrared temperature measurement; starting a black body, an infrared thermal image component, an industrial personal computer and a measuring software system; acquiring and displaying a thermal image video though an image acquiring card by the software system after the block body is constant at a certain temperature T, adjusting the parameter of the infrared thermal image component, and memorizing the thermal image in the form of bmp after the thermal image is clear; opening the memorized thermal image, selecting a black body region in the thermal image, and reading the mean gray level of the black body by the system; calling a database to search the calibrated gray level corresponding to the T temperature; and obtaining the error between the mean gray level and the calibrated gray level, wherein the error reflects the degree of the influence of the background factors to the infrared temperature measurement. The method can quantificationally measure the influence of the background factors to the infrared temperature measurement when practically measuring the temperature, wherein in result can be used in the compensation of the temperature measurement to finally improve the accuracy of the temperature measurement.

The invention discloses a coal rock fracture development infrared radiation monitoring test method which is suitable for a coal rock fracture development infrared radiation monitoring test for researching a coal rock deformation condition of a mine. Supports for arranging coal rock test blocks are respectively arranged on two sides of a workbench of a rock press; three prepared coal rock test blocks are respectively arranged on the workbench and the supports; the rock press applies pressure on the loaded coal rock test blocks according to set pressure and set speed; a computer is used for processing infrared radiation mean temperature data of the loaded coal rock test blocks and a reference coal rock test block so as to obtain real infrared radiation data required by coal rock fracture development. The shortcoming of high environment and background factor influence on the conventional coal rock test block loading destroy infrared radiation monitoring test is overcome; an error caused by test conditions and the environment is reduced, the accuracy, the scientificity and the effectiveness of a test result are greatly improved, and instruction significance for the coal rock test block infrared radiation monitoring test is realized.

The invention discloses an object surface emissivity field calibration method. The actual temperature value of a paint coating is measured through a platinum resistor patch, emissivity is calibrated, so that the platinum resistor patch serves as a calibration target; the calibration target is closely attached to a filed calibration object, temperature of local surface of the field calibration object with the calibration target as the center is measured through a thermal infrared imager, and the temperature of the surface of the field object and temperature data of the paint coating on the calibration target are obtained; the value of the surface emissivity of the field object is calculated according to the temperature of the surface of the field object and the temperature data of the paint coating on the calibration target, and field calibration is performed on the surface emissivity of the field object. According to the method, the temperature of the field calibration object is measured through a thermodetector, temperature measuring speed and accuracy are high, and the surface of the field calibration object and the calibration target are in the same plane so that background factors and system noise can have the same influence on the field calibration object and the calibration target; the method is suitable for emissivity calibration of the surface of the field object, and calibration equipment is convenient to carry.

The invention relates to a pedestrian clothing attribute recognition method based on depth attitude estimation and multi-feature fusion. The method comprises the following steps: firstly, selecting apart of retrieval results for subsequent attribute recognition through appearance feature matching; then, through an SSD-based depth human body posture estimation method, a foreground region belongingto a pedestrian in the image can be effectively positioned, and background factor interference can be well eliminated; and finally, fusing analysis results of multiple modes, and combining an iterative smoothing process to enhance the correlation between an attribute tag and a pixel by adopting a maximum posterior probability distribution mode to obtain a final attribute analysis recognition result. According to the method, the problems of inaccurate label recognition, pixel analysis area deviation and the like under a single analysis mode are solved. The method is simple and flexible, and has high practical applicability.

This invention discloses a multi-channel active controller, which comprises an error signal conditioner, a digital signal processor, a filter, a control signal driver and reference signal conditioner. Wherein, the signal from error sensor located outside the multi-channel active controller sequentially passes error signal conditioner, digital signal processor, filter and control signal driver; outer reference signal via reference signal condition is input to digital signal processor consisting of a plurality of A/D converters, real-time processor chips, and background factor updating data processor chips. The invention is operated simply, can maintain reality of time and use the controller traceability.

The invention discloses a near-surface scene infrared small target detection method, which comprises the following steps of (10) infrared image preprocessing: preprocessing an infrared image by adopting a hybrid filtering algorithm; (20) background suppression: performing background suppression on the preprocessed infrared image by adopting Robinson filtering to obtain a background suppression infrared image; (30) target enhancement: calculating an integral image of the background suppression image, calculating background factor parameters, and readjusting the image by using the background factor parameters to obtain a target enhanced image; (40) obtaining the centroid of the target potential area: carrying out threshold segmentation, morphological processing and connected domain marking on the target enhanced image to obtain the centroid coordinates of the target potential area; and (50) obtaining real coordinates of the target: performing time domain association and track discrimination on the potential target to obtain the real coordinates of the target. The small target detection method is high in robustness, low in false alarm rate and high in detection precision.

The invention relates to a plant leaf distinguishing method based on image segmentation and feature fusion. According to the method, the plant leaf identification in a complex background is divided into two stages of leaf segmentation and feature extraction. The morphological transform-based marker-driven watershed algorithm is mainly adopted for starting and stopping the reconstructing operationon plant leaves. After that, plant leaves are segmented by adopting the marker-driven watershed algorithm. As a result, the over-segmentation phenomenon caused by the influence of leaf textures and veins during the segmentation process of plant leaves is avoided. Therefore, the negative effect of the complex background factor on plant leaf identification is eliminated.

The invention discloses an unmanned aerial vehicle video multi-target tracking method based on attention feature fusion, and aims to solve the problems of interference, shielding, variable viewpoint height and angle and the like in an unmanned aerial vehicle multi-target tracking video due to complex background factors. An unmanned aerial vehicle multi-target tracking algorithm network model is constructed and trained, a backbone network of the network selects a 34-layer residual network, a feature extraction network combined with a triple attention mechanism is designed in a feature extraction part, and a cascade feature fusion module is designed in an up-sampling part; the multi-target expression ability is stronger due to optimized features brought by the attention mechanism designed by the invention, the small targets in the unmanned aerial vehicle aerial video can be tracked more favorably due to the designed multi-scale information fusion channel, and the accuracy of predicting the multi-target trajectory in the unmanned aerial vehicle video by the association algorithm is further improved due to the optimized features.

The invention relates to visual detection and recognition techniques and provides a deep-sensor-based equipment control system so that problems that the existing RGB/infrared-camera-based equipment control system has a privacy disclosure problem, the RGB-camera-based body positioning plan is affected by the color, illumination, and background factors easily, and the RGB camera without light supplementing can not work in a dark light or black environment can be solved. The equipment control system is characterized by comprising a deep data acquisition and storage module, a human body positioning module and a control module. The deep data acquisition and storage module is connected to the control module through the human body positioning module. The deep data acquisition and storage module is used for collecting and storing deep data and uses a deep sensor to collect data. The human body positioning module is used for acquiring body positioning information based on the deep data; and thecontrol module is used for controlling corresponding equipment based on the obtained body positioning information. The deep-sensor-based equipment control system has the following beneficial effects:the privacy of the user is protected; and the control system is suitable for intelligent home appliances.

The invention relates to a health risk assessment method and system for territorial space planning. The method comprises the steps of obtaining a spatial scale of an evaluation scheme; determining a grid unit for carrying out health risk assessment according to the spatial scale; acquiring health risk basic data; determining risk factor data and background factor data according to the health riskbasic data; performing health influence assignment on the risk factor data and the background factor data according to the grid units to obtain various risk factor influence analysis graphs and wind field simulation analysis graphs; performing spatial superposition analysis processing on the various risk factor influence analysis graphs and the wind field simulation analysis graph to obtain a health risk comprehensive analysis grid graph; and determining planning optimization suggestions according to the health risk comprehensive analysis grid map. The method can comprehensively evaluate the negative health influence of the territorial space current situation environment or planning scheme on residents.

A method, system and computer program product for determining the suitability of an entity to provide products or services. Category and measurement data is received concerning the entity where each category is assigned a value based on the acquisition context. If the value assigned to a category exceeds a threshold, then the measurement data for that category is used in evaluating the entity. This measurement data is weighted according to the acquisition context. A binary value is generated for each weighted measurement value that exceeds a threshold. These binary values are summed and weighted according to the confidence that the source of the data is correct. Further, the past performance and reputation of the entity is used in applying a weight to the summed binary values to generate a suitability value. If the suitability value exceeds a threshold, then it is deemed suitable to conduct business with the entity.

The invention discloses an image processing method and an image processing device. The image processing method is used for processing mixture of source image pixel points and background pixel points, and comprises the steps of: acquiring a source image alpha value, a source image global alpha value and a source image color value of the source image pixel points; selecting the source image alpha value, the source image global alpha value, or a product value of the source image alpha value and the source image global alpha value as a source image calculation value according to a source image blending mode value; selecting 0, 256, the source image calculation value, a difference value between 256 and the source image calculation value, or a background estimated value as a background factor value according to a source image blending factor value; and calculating a source picture color calculation value according to the source image color value and the source image calculation value. The image processing method integrates mix_alpha and port_duff alpha calculation algorithms, simplifies logic corresponding to the two different algorithms, reduces area required by alpha operating hardware, and improves power utilization rate.

The invention discloses a construction method of a prostate neuroendocrine carcinoma genetic engineering animal model. The construction method includes the following steps that a TRAMP mouse with the PSGL-1 gene knocked off is raised to 24 week old or more to obtain the prostate neuroendocrine carcinoma genetic engineering animal model. The model more naturally simulates the generation and development process of clinical diseases at the overall level, and complex biological problems and drug action are effectively combined to study the treatment effect of drugs. In addition, various background factors in the human body can be considered by carrying out study in the mouse model live animal constructed by use, and research conclusions are high in reliability. Therefore, it is found that the experiment mouse model is an effective tool for screening the drugs for resisting prostate neuroendocrine carcinoma and is of important significance in finding key targets of the effective treatment drugs and treatment methods of nerve diseases.

The invention belongs to the technical field of analysis and detection, and particularly provides a detection method of biomolecules and inhibitor molecules. The biomolecule concentration detection method comprises the following steps that: a to-be-detected substance and hydrogel are mixed to form a mixed solution, the mixed solution is placed one end of an indicating test strip, different amounts of target detection substances cause different hydrolysis degrees of the hydrogel, so that the content of water released by decomposition of the hydrogel is different, different distances of movement can be generated on the pH test strip, and the detection of the target detection object is realized through the length of the moving distance. The problems that a common biomolecule detection method in the prior art is tedious in operation steps, needs to train professionals to operate related equipment, is long in detection time and low in efficiency, causes resource waste and cost increase, and influences the detection sensitivity due to environment background factors such as matrixes exist are solved.

The invention discloses a crowd counting method combining density estimation and target detection. The method comprises the following steps: S1, marking training pictures, training a density estimation network, a target detection network and a picture partition network by using the training pictures, and constructing a fusion network; s2, inputting the picture into a fusion network, and obtaining a sparse region and a dense region of the picture according to a picture partition network; s3, using a density estimation network in the dense region of the picture to obtain a head density map; using a target detection network in the sparse region of the picture to obtain a head bounding box density map; and fusing the two density maps through a fusion network to obtain a fused density map, and carrying out crowd counting. According to the method provided by the invention, the monitoring image is processed and analyzed in real time, so that the method can adapt to characteristics of various different scales, the influence of background factors on counting is reduced, the counting error is reduced, and the accuracy of crowd counting is improved.

The invention provides a video transmission device, and the device comprises a preprocessing module which is disposed at a transmitting end, and is used for decomposing an original target image in a to-be-transmitted video into a foreground factor, a background factor, and an attitude factor; the encoding system is configured at the sending end and is used for encoding the original image based on the foreground factor, the background factor and the attitude factor to obtain foreground information embedding characteristics, background information embedding characteristics and attitude information embedding characteristics and transmitting the foreground information embedding characteristics, the background information embedding characteristics and the attitude information embedding characteristics to the decoding system; and the decoding system is configured at the receiving end and is used for decoding the received embedded features to generate a reconstructed image. According to the method, image data are deconstructed into a plurality of factors at a sending end, an input image is coded into an embedded feature based on each decomposition factor, then an image is reconstructed based on the embedded feature, and a more real image is generated by using a deep convolutional generative adversarial network. On the premise that the transmission bandwidth and storage resources are reduced, the quality of the generated image is improved, and the transmission quality is ensured.