31results about How to "Accurate aggregation" patented technology

The invention discloses a three-dimensional face expression recognition method based on an SSF-IL-CNN. According to the method, firstly, a convolutional neural network structure is improved, a convolutional kernel is disassembled into the structure parameters and the strength parameters, and the two parameters are enabled to respectively undertake the initialization and updating tasks, so that theconvolutional neural network originally needing a large number of samples for training can be applied to a three-dimensional facial expression library with a small sample scale. Meanwhile, accordingto the method, an Island Loss function is adopted to construct a loss function in the convolutional neural network, the sensitivity and the distinction degree of the network to the facial expressionsare enhanced, and the expression recognition effect is improved.

The invention discloses a network video event mining framework based on dynamic association rules. Aiming at the characteristics of little information, much noise and the like of network video text information, by using semantic information among keywords and by dynamically adjusting the support degree in the association rules, influences on the relation between text and visual information caused by problems of noise, synonyms, polysemous words, multiple languages and the like are reduced, the robustness of text information is enhanced, and a bridge between visually approximate key frames and high-level semantics is established. The network video event mining framework based on dynamic association rules fully utilizes the semantic correlation among the text information and the relation between the text information and the visual information, the influences of noises in the text information on the relation between the visually approximate key frames and events are reduced, the correlation between the visually approximate key frames and the corresponding events is enhanced, and the framework is enabled to better adapt to event mining of network videos with much noise and little information.

The invention discloses an infrared image volume cloud detection method based on neighborhood intensity texture coding. The invention relates to the technical field of infrared image processing. The method comprises the following steps: firstly, creating a sliding window to traverse infrared image pixel points; performing texture coding on the infrared image by using a neighborhood intensity coding method to generate a texture image, roughly determining coordinates of a suspected point region based on local probability distribution, then using region growth to aggregate false alarm source pixels, and finally performing denoising by using a neighborhood statistical method to finally obtain a detection result. According to the method, the problems of difficulty in accurately detecting the volume cloud, low detection precision and fuzzy detection result in the existing infrared imaging technology are solved, and the method has the advantage of accurately aggregating the false alarm sourcepixel region.

The invention discloses an infrared imaging river channel detection method based on neighborhood intensity texture coding. The invention relates to the technical field of infrared image processing. The method comprises the following steps of: performing texture coding on an input image by using a neighborhood intensity coding method; generating a texture image, roughly determining coordinates of asuspected point region based on local probability distribution, generating a preliminary detection result by using region growth, neighborhood statistical denoising and other methods, detecting a connected region of the preliminary detection result, taking gradient features of the input image, and performing region growth to finally obtain a detection result. According to the method, the problemsthat the riverway is difficult to accurately detect, the detection precision is low and the detection result is fuzzy in the existing infrared imaging technology are solved, and the method has the advantage of accurately aggregating the false alarm source pixel region.

The invention discloses a display device and a method for implementing holographic display thereof. A plurality of first lenses closely arranged in an array manner constitute a first lens array, and each of the first lenses corresponds to a plurality of observation regions arranged on a light outgoing path respectively; a plurality of slits arranged in an array manner and corresponding to the first lens one by one constitute a slit array; a spatial light modulator loads a holographic image; a human eye tracking device tracks the position of human eyes; a controller determines an observation region according to the position of the human eyes, then controls a deflection element to direct the exit light of the spatial light modulator to the slit corresponding to the first lens corresponding to the observation region, and controls the deflection angle of the first lens to converge the exit light of the first lens to the position of the human eyes. Because the controller can control the deflection of different angles of the first lenses according to the position of the human eyes, the holographic image can be accurately converged at the position of the human eyes, thereby overcoming theproblem that the viewing angle is limited in the holographic window technology.

The invention discloses a novel multi-channel handheld SPR detector. The novel multi-channel handheld SPR detector comprises a casing; a light incidence assembly, alight reflection assembly and a light reception assembly are sequentially clamped in a light incidence assembly accommodating cavity, a light reflection assembly accommodating cavity and a light reception assembly capacity; a sensing chip comprises a substrate and a gold-plated glass piece; sensing chip guiderails are symmetrically arranged on the inner surfaces of the front and rear sides of the casing; sensing chip brackets are glidingly arranged on the sensing chip guiderails; sensing chips are arranged in the sensing chip brackets; and injection port sealing and connection devices and extraction port sealing and connection devices are respectively arranged at positions, corresponding to a plurality of sample storage grooves and sample recovery grooves, on the surface of the casing. The novel multi-channel handheld SPR detector is capable of directly extracting a plurality of liquid samples from a sample storage vessel at one time without dripping the liquid samples into a loading slot after taking the sensing chips in exposed environment, thereby preventing the liquid samples from being polluted and then improving the detection correctness.

The invention relates to a household appliance (1) for the care of laundry items, having a collecting apparatus (5) for leakage water which is arranged on the base (4) of the household appliance (1), and a float (81) which is arranged in the collecting apparatus (5) and is configured for making electrical contact with a switch (8) as a function of the water level (W) in the collecting apparatus (5), wherein the float (81) is arranged in a first part region (5a) of the collecting apparatus (5), which first part region (5a) is separated from a second part region (5b) of the collecting apparatus (5) by a water damming wall (52).

The invention relates to a face recognition based control method of an exhibition hall robot arm. The control method is characterized in that a target face with the shortest physical distance to a lens of shooting equipment for the exhibition hall robot arm is accurately determined from a 360-degree panoramic image with a face feature algorithm, and then a target face is subjected to forward calculation by combining a face attribute algorithm, and attribute information of the target face is obtained; then, control materials such as height, actions and intonation adapted to the attribute information of the target face are generated for the exhibition hall robot arm, and the exhibition hall robot arm is correspondingly controlled according to the control materials. Thus, spatial pose changedisplay of the exhibition hall robot arm can be performed according to the control materials adapted to the attribute information of the target face when the presence of visitors in an exhibition hallarea is accurately identified, power consumption of the exhibition hall robot arm is reduced, and more people are attracted to the exhibition hall area; besides, accuracy of face detection can also be improved.

The invention discloses an auxiliary positioning and screw placement device for a proximal femur intramedullary nail, which comprises an arc-shaped positioning through cylinder, a femur trochanter positioning needle, a first screw, a second screw and a working channel assembly, wherein the arc-shaped positioning through cylinder is of a quarter arc structure; a positioning needle channel is transversely formed in one end of the arc-shaped positioning through cylinder, the femoral trochanter positioning needle is movably inserted into the positioning needle channel through a first screw, and the needle point of the femoral trochanter positioning needle faces the circle center of the arc-shaped positioning through cylinder; the other end of the arc-shaped positioning through cylinder is provided with a threaded hole, the second screw is movably inserted into the threaded hole in the other end of the arc-shaped positioning through cylinder, and the working channel assembly is movably arranged at the other end of the arc-shaped positioning through cylinder through the second screw. The auxiliary positioning and nail placing device for the proximal femoral nail is simple in structure, accurate positioning can be achieved, and the perspective frequency is reduced.

The invention relates to a high-speed full-automatic grouping and track combining machine device. The high-speed full-automatic grouping and track combining machine device is characterized by comprising a first grouping mechanism, a second grouping mechanism and a material combining mechanism; the first grouping mechanism comprises a first material guide frame, a first chain belt, a first drivingmotor, a first bearing belt and a plurality of first material distributing components; the first material guide frame is provided with a first material guide channel; the second grouping mechanism comprises a second material guide frame, a second chain belt, a second driving motor, a second supporting belt and a plurality of second material distributing components; the second material guide frameis provided with a second material guide channel; the material combining mechanism comprises a conveying belt, a first guide piece, a second guide piece and a third driving motor; the input end of theconveying belt is simultaneously connected with the first material guide channel and the second material guide channel. Grabbing and placing beverage bottles by workers can be completely avoided, thepackaging efficiency can be greatly improved by the high-speed full-automatic grouping and track combining machine device, the situation of packing errors can be avoided, the hygiene of the beveragebottles can be prevented from being affected, and the high-speed full-automatic grouping and track combining machine device is very reliable in overall structure.

The invention provides a medical data aggregation method and a device, and the method comprises the steps: obtaining a to-be-aggregated business table which comprises at least one field, each field comprises at least one piece of medical data, wherein the number of medical data in the to-be-aggregated business table is greater than a first preset number; randomly collecting a second preset numberof samples in the medical data; carrying out first identification on the collected sample to obtain a first sample and a second sample, and determining a first type of a field corresponding to the first sample; performing second identification on the second sample, and determining a second type of a field corresponding to the second sample; and according to the first type and the second type, determining a service table category of the to-be-aggregated business table. According to the scheme of the invention, the medical data can be gathered more accurately.

The invention discloses a magnetic steering targeted therapy device which comprises a base, a vertical main lifting rod, a transverse telescopic rod, a vertical auxiliary lifting rod and magnetic generation plates. The base is connected with the lower end of the vertical main lifting rod, the upper end of the vertical main lifting rod is connected with the left end of the transverse telescopic rod, the right end of the transverse telescopic rod is connected with the vertical auxiliary lifting rod, and the two ends of the vertical auxiliary lifting rod are each connected with one magnetic generation plate. The device is characterized in that a central controller is fixedly connected to the joint of the transverse telescopic rod and the vertical main lifting rod, and a transformer, a microprocessor chip and a memory are integrated and connected in the central controller; the right end of the transverse telescopic rod is connected with the middle of the vertical auxiliary lifting rod through a full-angle rotating shaft; the upper section and the lower section of the vertical auxiliary lifting rod can ascend and descend; and the distance between the two magnetic generation plates and the body surface and the magnetic force can be automatically controlled and adjusted, and through the series of automatic control and adjustment, medicine can be completely and accurately attracted to the focus position to act.