35results about How to "To achieve the effect of real-time detection" patented technology

The detection method establishes dialog based on reliability detection mechanism (RDM) between two network devices (ND). The method includes steps: link state on ND is bound on the established dialog based on RDM; ND detects link state bound on this end, and tells the link state to ND at opposite terminal through dialog based on RDM. The invention also discloses means for implementing rearrangement of route strategy. The means includes steps: receiving link state, ND configures coincidence relation between route strategy and link; ND determines link corresponding to selected route strategy currently, and carries out rearrangement of route strategy based on the link state. The invention also discloses link detection device of implementing end to end, as well as device for implementing rearrangement of route strategy. The invention can realize link detection from end to end, and trigger off rearrangement in time.

The invention discloses a real-time detection method for road vehicles based on a deep learning SSD framework, which mainly solves the problem of low detection accuracy in the prior art under conditions of crowded roads and complex weather conditions. The implementation plan is: 1. Take several videos of driving vehicles in the traffic arteries, and capture the window information of the vehicles in each frame of the video by manual labeling; 2. Use VGG‑16 in the classification network as the basic network , build the SSD300*300 detection framework, extract the features of some of the feature extraction layers and connect them as the recognition feature vector, and input them into the loss function; 3. Use the training samples for training, and substitute the trained detection model into the detection framework Middle; 4. Set the correlation threshold and use the trained detection model to detect the vehicles in the test video. The invention not only greatly improves the detection accuracy, but also achieves the effect of real-time detection, and can be used for vehicle detection in complex scenes.

The present invention provides a ship detection method and system combining saliency detection and deep learning, including building a ship image sample library, including collecting monitoring video data of coastal areas under visible light, obtaining frame images containing ships, and then obtaining ship positions and The true value of length and width; construct a YOLO-like convolutional neural network, perform model training on the video ship target sample, and obtain the training result model of the ship target under the surveillance video; input ship image data, use the training result model to predict the detection bounding box; use the boundary Box information is used for saliency detection to obtain more accurate ship positions. The technical scheme of the invention is very fast and efficient, and can achieve the effect of real-time detection. It can also have good detection results for complex scenes such as clouds, cloudy days, and rain, and the method has high robustness.

The invention discloses a high-precision real-time multi-scale pointer detection method, which belongs to the image processing technology. The method of the present invention marks the dial picture in the training set, records the position of the pointer area in the dial picture; then performs data enhancement operation on the dial picture in the training set; then inputs the training set to the FaceBoxes network, and trains to obtain the dial pointer detection model; finally Input the image to be tested to the dial pointer detection model, output the pointer frame in the image to be tested, and then determine the pointer pointing by the center of the dial and the center point of the pointer frame; at the same time, use the adaptive Hough line detection for the detected pointer frame to further Refining the pointer pointing, combining the above two pointer pointing extraction methods, giving the preferred pointing information of the pointer; the invention also discloses a high-precision real-time multi-scale dial pointer detection system, which not only greatly improves the efficiency of pointer detection and accuracy, and the algorithm is more stable and robust.

The invention discloses a receiving antenna diversity fault detection method and device. The receiving antenna diversity fault detection method comprises the following step: calculating the signal intensity of each diversity for many times in a preset time, selecting a plurality of groups of signal intensity as effective detection data according to a pre-scheduled section principle, detecting the fault or fault recovery of each diversity for many times according to the plurality of groups of effective detection data, and determining and reporting that the diversity is in the fault state or fault recovery state when the times of detecting the fault or fault recovery of each diversity reaches the pre-scheduled condition. The receiving antenna diversity fault detection method and device overcome the problem that the diversity receiving abnormity can not be detected completely effectively in the prior art, are easy to implement, achieve the real-time detection effect, save the hardware detection circuit cost and improve the detection accuracy.

The invention disclose a digital electronic shear speckle interferometer based on a micro-polaroid array. The digital electronic shear speckle interferometer comprises a laser, a beam expander located on a light path of an outgoing beam of the laser, a to-be-detected object, a shearer located on a light path of a beam reflected from the surface of the to-be-detected object and used for obtaining two mutually perpendicular coherent linearly polarized lights, the micro-polaroid array arranged on a light path of a beam outgoing from the shearer, a photosensitive element connected with the micro-polaroid array, and processing equipment connected with the photosensitive element and used for transmitting and processing a signal acquired by the photosensitive element. The interferometer has the advantages that multiple phase shifts can be obtained by single exposure; a real-time detection effect can be achieved; a phase shift device is simplified in the light path; and as the micro-polaroid array generating the phase shifts is directly integrated onto the photosensitive element, the antivibration requirement is lowered.

The invention discloses a method for detecting abnormal events of a network host based on a mobile agent. According to the detection result of a monitoring system, a host data collection and analysis mobile agent is dispatched to a sensitive host to perform data collection and security analysis tasks. Firstly, the mobile agent collects host resource information including CPU utilization, hard disk IO, network traffic, system process, memory utilization and other host resource information in the target host. The black and white list method is used to judge whether there are illegal processes, the NetFlow model is used to extract traffic characteristics and compared with the normal mode to judge whether there is abnormal traffic, and the multi-source information fusion method is used to fuse and analyze host information to judge whether there is abnormality. According to the analysis results, a dynamic isolation control strategy is adopted for abnormal hosts to reduce their security threats to other network hosts. The host data collection and abnormal behavior detection method proposed in this patent is simple and efficient, the data collection and analysis tasks are small, and it can be applied online in real time.

The present invention relates to the technical field of seawater detection, in particular to a detection and alarm device for mariculture, comprising a mounting plate, the upper end of the mounting plate is provided with a motor, the output shaft of the motor is mated and connected to the winch, the mounting plate The lower end of the plate is also fixedly connected with a vertical rod, the side end of the vertical rod is provided with a chute, and a sliding rod is slidably connected to the chute, and one end of the sliding rod is fixedly connected with a detection box. If the limit rod touches the edge of the upper end of the card hole when it moves downward, it will guide the second limit rod, thereby improving the practicality of the equipment, avoiding the slipping of the belt, and accurately controlling the detection box The depth in the pool, and the detection box can work stably in the pool through the limit component, which improves the practicability of the equipment, and can detect the temperature and pH value of the seawater inside the breeding pool in real time, providing timely and accurate information.

The invention discloses a sidewalk vehicle illegal parking detection method based on target detection and semantic segmentation, which belongs to the field of deep learning and computer vision. In a training stage, a semantic segmentation network and a target detection network are respectively trained to obtain a semantic segmentation model and a target detection model; In the testing stage, the mixed Gaussian background model is used to extract the urban road background map, and then the semantic segmentation algorithm and the semantic segmentation model are used to perform semantic segmentation to obtain the semantic segmentation map. Then, the perceptual hash algorithm is used to determine whether the camera is rotating, and then the target detection is used. The algorithm and the target detection model perform vehicle detection, mark the vehicle detection frame, and finally compare the vehicle detection frame with the semantic segmentation map. The invention solves the problem that the existing vehicle illegal parking detection method has a high misjudgment rate, poor real-time performance, needs to manually demarcate the illegal parking area on the sidewalk, and when the camera rotates, it needs to manually demarcate the illegal parking area, which is not suitable for rotating the camera. The problem.