74results about How to "Reduce the probability of false detection" patented technology

The invention discloses a high reflection surface defect detection method based on image processing and neural network classification. The method includes: performing basic processing including background deduction and denoising on an originally acquired image, preliminarily determining positions of parts possibly having defects through feature extraction and obtaining a series of area images of the corresponding positions, inputting an image sequence to a neural network classifier regarding a local feature area block as training data, determining whether the defects are real defects, and regarding an output result of the classifier as a final determination result. According to the high reflection surface defect detection method, defect feature search and extraction are performed by employing a front-end two-dimensional digital image processing module, feature filtering and enhancement are performed with the combination of a rear-end neural network classifier, the accuracy of the search result is enhanced, surface defects of a detected member are fully extracted, the probability of false detection and omitted detection in the conventional image processing detection method is reduced, and the operation efficiency and the versatility are simultaneously considered.

The invention discloses a method for human face authentication. The method comprises the following steps: the fisher coefficient of a target to be authenticated is calculated with the fisher eigenface method based on principal component analysis; the similarity matching degree corresponding to the fisher coefficient is calculated by utilizing a lookup table classifier; the similarity matching degree and a preset threshold are compared to get the authentication result. The invention also discloses a device for human face authentication. The device comprises a coefficient computing module, which computes the fisher coefficient of a target to be authenticated; an output computing module, which computes the similarity matching degree corresponding to the fisher coefficient with a lookup table classifier; and an authenticating module, which compares the similarity matching degree and the preset threshold to get the authentication result. The method and the device for the human face authentication of the invention, by introducing the lookup table classifier with more powerful expressing ability than a single threshold, can perform better classification on training samples, thereby improving the recognition rate of human face authentication and reducing the mis-detection possibility.

The invention relates to a detection method for building change by a multispectral image. The detection method comprises the step of detecting the change based on a pixel ratio method to obtain change regions of all ground objects firstly, wherein the change regions are used as a change candidate region of the building. The change detection based on building feature is carried out by using the change regions detected by using a ratio method. False detection easily occurs by adopting one feature, so according to the method, feature change detection is carried out by adopting a mode of sequentially combining texture feature with tone feature to distinguish the change of the building from the change of other ground objects, and thus detection accuracy of the building is increased and the false detection rate is reduced.

The invention provides a character area image acquiring method, including: executing letter testing to at least two frames of target images, to obtain the candidate character area image on each frame of target image; executing inter-comparison process between the candidate character area image on one frame of target image with the candidate character area image on the other frames of target images that may contain the identical character content; if the difference between the inter-compared at least two candidate character area images is in the prearranged scope, determining that the at least two candidate character area images pertain to the character area image including the character content; and acquiring the character area images. The invention also provides a character area image acquiring method and a character recognition system. The technical scheme provided by the embodiment of the invention has passed the verification to the character testing result, and reduces the miss-detection probability of the character testing.

The invention discloses a method and a device for judging an error detection of a physical downlink control channel. The method comprises the following steps: making a hard decision on a soft-bit data output after a rate de-matching; decoding the soft-bit data and carrying out cyclic redundancy check, coding the decoded data after the cyclic redundancy check; comparing the coded data with the result of the hard decision, calculating the ratios of different bits between the coded data and the result of the hard decision to the soft-bit data; judging whether the ratio is greater than a scheduled threshold value or not, if so, judging to be error detection. According to the technical scheme, the problem that PDCCH (physical downlink control channel) error detection can not occur under an LTE (long term evolution) system can be solved, and the probability of the occurrence of the PDCCH error detection is greatly reduced.

The invention relates to the field of situation data fusion of the wireless sensor network, particularly to trust-based situation data fusion method of a distributed sensor network. An event fusion center receives event occurrence states detected by all association sensing nodes, determines whether events occur or not, and adjusts credibility of the association sensing nodes; fusion centers carry out fusion processing on event data detected by all association sensing nodes; and a base station receives fusion values of all fusion centers, carries out the second fusion to obtain secondary fusion numerical values, detects consistency of the fusion centers, and issues a final fusion data result. According to the invention, steps of event detection, situation data fusion and consistency detection are carried out. Fusion weights are distributed based on credibility of the sensing nodes and the historical credibility, the time relevancy, and the data relevancy are taken into consideration. Therefore, phenomena of information concealing and false information providing can be effectively prevented; the wrong detection probability is reduced; a network performance in a state with large occupation ratio of the normal sensing node number can be improved; the fusion value is close to the actual value; and the data accuracy and reliability are improved.

The invention discloses a rectangular seal positioning method and a device aiming at a certificate image. The method comprises the following steps of: S1, sequentially judging whether a color of each pixel in an original image is the color of a specified seal, and generating a binary image of original image; S2, generating an edge binary image based on the binary image; S3, setting a window, judging whether a total amount of a foreground ponit in the window in the edge binary image is less than a preset noise threshold, if so, moving the window to rejudge in the edge binary image, if not, calculating a center of gravity of the foreground ponit, and determining whether the centre of gravity is within a preset area, if so, determining that positioning is carried out successfully, and recording the edge binary image in the window, if not, moving the window in the edge binary image and rejudgeing. According to the method and device, some characteristics of a seal image are considered, a foreground image with a specified color is extracted, and the window is set to slid to intercept a rectangle to initially judge a position of the rectangle in the whole image, so that rectangle detection aiming at the seal image in a later period can be carried out within a small scope, and probability of false detection and a calculated amount are reduced.

The invention relates to the technical field of communication, and discloses a method and a device for suppressing false detection of a random access channel. The method comprises the steps of acquiring a first position of a first frequency domain resource allocated by a physical random access channel (PRACH) based on configuration information of the PRACH; determining physical uplink shared channel (PUSCH) service scheduling information required to be scheduled by an up link (UL) subframe at the next moment, judging historical signal to noise ratio (SINR) information of each user equipment (UE) of a scheduled PUSCH service within a set period of time before the current moment according to the service scheduling information, and determining an SINR average value of each UE according to the SINR information; and allocating second frequency domain resource for each UE according to the SINR average value corresponding to each UE and the first position, wherein when the SINR average value corresponding to the first UE is greater than the SINR average value corresponding to the second UE, the distance between the position of a frequency domain resource which is allocated to the first UE correspondingly and the first position is greater than the distance between the position of a frequency domain resource which is allocated to the second UE correspondingly and the first position.

The invention relates to a blind detection method for a control channel, a terminal, a network element and a storage medium, and the method comprises the steps: obtaining downlink control informationDCI of a plurality of subframes through the blind detection of the control channel; and within the repeated sending times of the control channel, performing validity judgment on the DCI of the plurality of subframes to obtain a judgment result. According to the embodiment of the invention, the effectiveness judgment is carried out on the DCI of the plurality of subframes obtained by blind detection within the repeated sending times of the control channel to obtain the judgment result of whether the DCI is the false detection or not, so that the false detection probability is reduced, and the blind detection accuracy of the control channel is improved.

The invention provides a panel lightspot detection method and a system, used for detecting the lightspot defect of a panel. According to the method, an image acquiring module acquires a transmitted image on behalf of a lightspot defect and a foreign matter. Meanwhile, the image acquiring module acquires a reflected image. In the reflected image, the part of a higher brightness is formed through a polarizer when the light is reflected by the foreign matter on the panel. The reflected light is in the form of the linearly polarized light polarized along a first direction when no foreign matter exists on the panel. The polarizer is provided with an optical axis and the optical axis is vertical to the first direction so as to block the linearly polarized light. A processing module conducts the image subtraction operation. That means, all pixel values of the transmitted image are subtracted by all pixel values of the reflected image to form a result image. In the result image, the part of the higher brightness is just the location of the lightspot defect.

The invention discloses a testing method for radio frequency of a mobile phone. The mobile phone is registered to a 3G network, and a radio frequency index of the 3G network is tested; if a test result is qualified, the 3G network is switched into a 2G network; and if the 3G network is switched into the 2G network successfully, the radio frequency index of the 2G network is tested, if a test result is qualified, the test ends and PASS is displayed. If the test results of the radio frequency indexes of the 3G network and the 2G network are failed, the radio frequency indexes are tested repeatedly for many times, until fail times reach a predetermined value, FALL is displayed. By testing for many times, the method effectively reduces mis-test probability, improves testing efficiency and ensures communication quality.

The invention discloses an efficient detector used for docking stations. The efficient detector includes a pedestal and a dividing support dividing a space above the pedestal into an upper area and alower area. A first detection unit is arranged in the upper area and a second detection unit is arranged in the lower area. Each of the first detection and the second detection unit includes a first support rack, a first drive device, a second support rack, a third support rack and a fourth support rack, wherein the first support rack is provided with a first positioning piece, the second is provided with a second drive device and a second positioning piece, the third support rack is provided with a third drive device and the fourth support rack is provided with a detection assembly. The top end of the dividing support is also provided with two fifth support racks arranged in a staggered manner. Each of the fifth support racks is equipped with a second image collector. The detector provided by the invention can perform detection of two docking stations at the same time; the detection efficiency is improved and the detection efficiency can also be ensured.

The invention provides a pin hole flaw detection system for detecting isolated film, which comprises the following steps: an optical camera device on an isolated film to be detected containing points to be measured captures detection images of the points to be measured; light source provides light moves towards the optical camera device; a first polarization component is arranged between the light source and the optical camera device; a second polarization component is arranged between the first polarization component and the optical camera device; a control module is coupled with the optical camera device for processing detection images; wherein, the isolated film to be detected is arranged between the first polarization component and the second polarization component, and the included angle between a first optical axis of the first polarization component and a second optical axis of the second polarization component is 90 degree for capturing detection images; the pin hole flaw detection system of the present invention can effectively detect pin hole flaws on the isolated film, thereby minimizing false detection probability and avoiding blast caused by battery short circuit.

The invention discloses a pavement water body detection method which comprises the following steps: acquiring a road scene image by using a visible light camera, zooming the road scene image to a specified size, and acquiring a mask containing water body position information in the acquired image by using a labeling method; constructing a conditional generative adversarial network combined with areflection attention unit, and training the conditional generative adversarial network combined with the reflection attention unit by utilizing the acquired road scene image and the marked mask; and zooming an image to be detected to a specified size, inputting the scaled image into the trained conditional generative adversarial network combined with the attention reflection unit, and obtaining abinary image which is output by a generator of the conditional generative adversarial network and is used for representing a water body detection result. According to the pavement water body detectionmethod, the surface water body area can be accurately and efficiently detected by utilizing the pavement image acquired by the camera, and the detection result has relatively high accuracy and recallrate, and the pavement water body detection method can be applied to tasks related to surface classification in the field of unmanned driving.

The embodiment of the invention discloses a voice signal processing method and device, and a terminal. The method comprises the steps: obtaining a to-be-processed voice signal and at least two reference signals; calculating cross-correlation parameters of the to-be-processed voice signal and the at least two reference signals; and if it is determined that the to-be-processed voice signal has a target voice signal according to the cross-correlation parameterS, performing dynamic range compression (DRC) processing on the to-be-processed voice signal. According to the technical scheme of the embodiment of the invention, the performance of voice signal DRC processing can be improved, so that the false detection probability is reduced, and the user experience performance is improved.

The invention discloses an interference object detection method and equipment. The method comprises steps that: first pixel points of a present frame image are contrasted with second pixel points of a reference frame image to determine the first pixel points and the second pixel points which satisfy conditions, and a first tonal value is acquired according to the first pixel points; third pixel points are searched from the present frame image, and a third tonal value is acquired; a second tonal value in the reference frame image is acquired through the second pixel points, a first difference value of the first tonal value and the second tonal value is respectively calculated, a second difference value of the first tonal value and the third tonal value is respectively calculated, if the first difference value and the second difference value are both smaller than a first set threshold, the first pixel points is determined to be pixel points of an interference object, and the interference object is detected according to the determined pixel points of the interference object. Compared with the prior art, a calculation method is relatively simple through a point mode, and accuracy for determining the interference object is relatively high.

The invention discloses a false detection filtering method and device, a storage medium and a receiver. The method comprises the following steps: decoding and verifying to-be-processed data; if the verification is passed, encoding the decoded data obtained by decoding to obtain encoded data; comparing the encoded data with the to-be-processed data bit by bit, and counting the number of error bits;and determining whether a decoding verification result of the to-be-processed data is false detection or not according to the number of the error bits. Through the scheme of the invention, the falsedetection probability can be effectively reduced.

The embodiment of the invention provides a detection method and a detection device which are applied to a narrowband Internet of Things base station, and the method comprises the following steps: receiving an uplink signal from user equipment; determining signal effective power corresponding to the uplink signal and a signal-to-noise ratio corresponding to the uplink signal; and if the signal-to-noise ratio exceeds a first threshold, or if the effective power of the signal exceeds a second threshold, determining that the uplink signal comprises a random access lead code. The first threshold and / or the second threshold are / is determined according to the repeated sending times of the random access preamble corresponding to the user equipment. According to the embodiment of the invention, thefalse detection probability and the missing detection probability of the Preamble code can be reduced, the accuracy of detecting the Preamble code is improved, and the efficiency of accessing the NB-IoT by the user equipment in a scene of high-speed movement can be further improved.

A method of detecting the states of power cables in an inverter system supplying power generated from an inverter to a motor by using three phase power cables is provided. The method includes: calculating the location of a current space vector for a first period when the first period arrives; using the calculated location of the current space vector for the first period to calculate the predicted location of the current space vector for a second period; calculating the actual location of the current space vector for the second period when the second period arrives; comparing the calculated predicted location with the actual location; and detecting the states of the three phase power cables according to a comparison result.

The invention discloses a steel pipe surface defect detection device and method and relates to the technical field of pipe detection. The steel pipe surface defect detection device comprises a steel pipe carrier roller mechanism, a steel pipe, a camera movement control mechanism, a camera, a steel pipe clamping and rotating mechanism and a main control computer. A steel pipe surface defect detection method is applied to a steel pipe surface defect detection device and comprises the following steps of 1, presetting items; 2, segmenting the steel pipe; 3, clamping the steel pipe; 4, acquiring a steel pipe surface image; and 5, comparing the images. The steel pipe surface quality detection device can replace manual work for steel pipe surface quality detection and is suitable for surface quality detection of different pipe diameters, different lengths and different detection precision requirements. Compared with manual detection, detection efficiency is higher, and the error and leakage detection rate is greatly reduced.

The invention discloses a target extraction method based on ground monitoring radar spectrum data. The target extraction method comprises the following steps that spectrum data information processed by a radar front end is obtained; the spectrum data information is subjected to cancellation processing of spectrum data to obtain spectrum data information after cancellation processing; constant false alarm rate processing is carried out on the spectrum data information after cancellation processing to obtain spectrum data information of a target, and a two-dimensional image is constructed according to the spectrum data information of the target; target edge extraction is carried out on the two-dimensional image to extract spectrum data information belonging to the same target; and ground object target filtering is carried out on the spectrum data information belonging to the same target to obtain spectrum data information belonging to the same target after the ground object target filtering. The target extraction method reduces the influence of fixed strong clutter on low-speed targets, improves the detection probability of the low-speed targets, and provides a reliable basis for subsequent target cluster and track processing.

The invention discloses a method and a device for frame synchronization of interphone communication. The method comprises the steps of carrying out correlation operation for a to-be-measured sequence extracted in a received signal and a preset normalized synchronous sequence to obtain a correlation value; calculating a normalization factor by utilizing an element with the absolute value thereof in the to-be-measured sequence satisfying a set range, wherein the set range is greater than 2; carrying out normalization processing for the correlation value by utilizing the normalization factor to obtain a normalization value; and if the normalization value is greater than a set threshold value, determining that the to-be-measured sequence is a synchronous sequence. The invention also discloses a device for frame synchronization of interphone communication, which adopts the method abovementioned. With adoption of the method and the device for frame synchronization of interphone communication, false detection of frame synchronization caused by noise in an interphone communication process can be avoided, the false detection probability is reduced efficiently, the reliability of frame synchronization is improved, and the communication quality of an interphone is further improved.

The invention discloses a pavement obstacle detection method based on reverse feature matching. Remote sensing change detection field, the method is based on space-air high-resolution remote sensing images. Change information in an image is extracted by using an SIFT feature point reverse matching method, the area and position information of a pavement obstacle in the image is obtained by processing the change information, the method has the characteristics of large-scale and macroscopic detection, and reference information can be provided for emergency rescue and disaster relief route planning and road rush-repair engineering quantity estimation. The method is characterized in that the method comprises the steps that SIFT feature point sets are calculated for remote sensing images beforeand after road loss respectively, feature matching is carried out, and negation is carried out on a matching result to obtain a change information feature point set; screening by using a road buffer area to obtain a change information feature point set in the road area, and screening out false change points and tiny change points by using a sliding window method; and finally, grouping and combining the feature points to obtain position information of the obstacle, obtaining an approximate range of the obstacle in combination with a region growing method, and performing area estimation. The keytechnology of the invention belongs to the field of remote sensing change detection, but the position and area information of the road obstacle can be obtained without image registration, and the degree of automation is high.

The invention provides a character area image acquiring method, including: executing letter testing to at least two frames of target images, to obtain the candidate character area image on each frame of target image; executing inter-comparison process between the candidate character area image on one frame of target image with the candidate character area image on the other frames of target images that may contain the identical character content; if the difference between the inter-compared at least two candidate character area images is in the prearranged scope, determining that the at leasttwo candidate character area images pertain to the character area image including the character content; and acquiring the character area images. The invention also provides a character area image acquiring method and a character recognition system. The technical scheme provided by the embodiment of the invention has passed the verification to the character testing result, and reduces the miss-detection probability of the character testing.

The invention belongs to the field of automatic production assembly machines and particularly relates to a laminating spinning connection device. The laminating spinning connection device comprises an upper mold and a lower mold; the lower mold is provided with a screw spinning connection mechanism; the screw spinning connection mechanism comprises a first telescopic device, a first sensor and a controller; a fixed part of the first telescopic device is fixed to the lower mold, and a telescopic part of the first telescopic device is provided with a plurality of drive devices through a connecting plate; output shafts of the drive devices are parallel to the telescopic direction of the first telescopic device and are connected with sleeves for positioning and storing adjusting screws; the first sensor is fixed to the lower mold through a mounting support, faces the plate surface of the connecting plate and is electrically connected with the input end of the controller; and the output end of the controller is electrically connected with the first telescopic device and the drive devices. The laminating spinning connection device has the beneficial effects that spinning connection of the two adjusting screws can be achieved, and the adjusting screws do not need to be screwed into a shell manually through a screwdriver, so that the labor intensity is relieved, and the automation degree, the production efficiency and the mounting quality are improved.

The embodiment of the invention provides a method and device for detecting a preamble signal in the technical field of communications. The method comprises the following steps: demodulating a receivedto-be-detected signal, determining a channel estimation value of each symbol group in the to-be-detected signal according to a preset frequency hopping characteristic of the preamble signal, determining effective power of the to-be-detected signal and a power correction value of the to-be-detected signal according to the frequency hopping characteristic and the channel estimation value of each symbol group in the to-be-detected signal; correcting the effective power of the to-be-detected signal by adopting the power correction value of the to-be-detected signal, determining the correction power of the to-be-detected signal, and finally determining whether the to-be-detected signal is a leader signal according to the correction power of the to-be-detected signal. The power correction valueis determined by combining the influence of the frequency offset on the effective power, and then the effective power is corrected by adopting the power correction value, so that the influence of thefrequency offset can be effectively reduced when the correction power is adopted to judge whether the to-be-detected signal is the preamble signal or not, and the omission ratio and the false detection rate of the preamble signal are reduced.

The invention discloses a high-reflection surface defect detection method based on image processing and neural network classification. The original collected image is first subjected to basic processing such as background subtraction and denoising, and the parts that may have defects are preliminarily determined through feature extraction. A series of regional images of corresponding parts are obtained, and then the image sequence is input into the neural network classifier obtained by using the local feature area tiles as training data to judge whether it is a true defect, and the output result of the classifier is used as the final judgment result. The invention proposes a surface defect detection method of a highly reflective test piece, which uses the front-end two-dimensional digital image processing module to search and extract defect features, and combines the back-end neural network classifier for feature filtering to enhance the accuracy of search results. Under the premise of fully extracting the surface defects of the test piece, the probability of wrong and missed detection in the traditional image processing detection method can be reduced, and at the same time, the computing efficiency and versatility can be taken into account.