Self-adaptive cascade classifier training method based on online learning

The invention discloses a self-adaptive cascade classifier training method based on online learning, which comprises the following steps: (1), preparing a training sample set with a small quantity of samples, and training an initial cascade classifier HC(x) in a cascade classifier algorithm; (2), using the HC(x) for traversal of image frames to be detected, extracting areas with sizes identical to the sizes of the training samples one by one, calculating a feature value set, classifying the areas with the initial cascade classifier, and judging whether the areas are target areas, thereby completing target detection; (3) tracking the detected targets in a particle filtering algorithm, verifying the target detection results through tracking, marking detection with errors as a negative sample for online learning, obtaining different attitudes of real targets through tracking and extracting a positive sample for online learning; and (4) carrying out online training and updating for the initial ...

Method for counting passenger flow of buses in real time

The invention provides a method for counting passenger flow of buses in real time, which adopts means of machine learning, multi-target detection and tracking, target behavior analysis and the like, and belongs to the technical field of pattern recognition. The method comprises the following concrete steps: detecting passenger targets by analyzing shape information and texture information of passenger heads in video images, wherein a column diagram in a gradient direction is used for representing the shape information, and a column diagram in a local binary mode is used for representing the texture information; then, accurately positioning passengers by adopting the target tracking policy combining gray level cross-correlation matching tracking and mean translation algorithm searching tracking; and finally, judging the behavior characteristics of the passengers by analyzing the moving tracks of the passengers, thereby accurately counting the passenger flow of buses. The practice shows that ...

Semi-supervision and classification method for hyper-spectral remote sensing images based on local stream type learning composition

The invention discloses a semi-supervision and classification method for hyper-spectral remote sensing images based on local stream type learning composition. The method comprises the steps: (1) preparing a training sample set, including a small amount of marked data and a large number of non-marked data; (2) choosing k nearest neighbor points for each sample point in the training sample set based on the distance measurement method of spectral angle mapping; utilizing a local stream type learning algorithm to obtain weights among connecting points in a graphic structure and calculating a graphic adjacency matrix to obtain the corresponding graphic structure; classifying the non-marked data based on the graphic adjacency matrix and a GFHF algorithm; and (3) classifying other data points in the images by using a GFHF generalized algorithm. Two widely applied algorithms, including a local stream type learning dimension-reduction algorithm and a semi-supervision and classification algorithm ...

Semi-autonomous on-line studying method based on random fern classifier

The invention provides a semi-autonomous on-line studying method based on a random fern classifier. The semi-autonomous on-line studying method based on the random fern classifier comprises the following steps that (1) a sample set for initial training of the classifier is prepared; (2) initial training of the random fern classifier is carried out; (3) on-line study samples are obtained; (4) on-line training of the random fern classifier is carried out. By means of the semi-autonomous on-line studying method based on the random fern classifier, the performance of a target detection system can be improved gradually, and the requirement for the detection accuracy is finally met. In addition, correctness of category marking of the on-line study samples is guaranteed through the method of obtaining the on-line study sample through manual judgment. Compared with a traditional classifier training method, the semi-autonomous on-line studying method based on the random fern classifier does not ...

Variable-frequency hydraulic oscillator

The invention relates to a variable-frequency hydraulic oscillator, and belongs to the technical field of oil exploitation downhole tools. The hydraulic oscillator is composed of a power short section and an oscillation short section, the oscillation short section is connected to the power short section, the power short section is composed of a power shell, a movable valve disc and a static valve, an upper power connector and a lower power connector are installed on the power shell in a threaded mode, and the movable valve disc is installed in the power shell through a screw and the movable valve connector. And a static valve is arranged between the movable valve disc and the power upper joint. According to the hydraulic oscillator, the flow area of the eccentric hole of the movable valve disc and the center hole of the adjusting plate can be periodically changed, then the pressure of drilling fluid is periodically changed, and therefore the piston is matched with the disc spring to drive ...

Self-adaptive cascade classifier training method based on online learning

The invention discloses a self-adaptive cascade classifier training method based on online learning, which comprises the following steps: (1), preparing a training sample set with a small quantity of samples, and training an initial cascade classifier HC(x) in a cascade classifier algorithm; (2), using the HC(x) for traversal of image frames to be detected, extracting areas with sizes identical to the sizes of the training samples one by one, calculating a feature value set, classifying the areas with the initial cascade classifier, and judging whether the areas are target areas, thereby completing target detection; (3) tracking the detected targets in a particle filtering algorithm, verifying the target detection results through tracking, marking detection with errors as a negative sample for online learning, obtaining different attitudes of real targets through tracking and extracting a positive sample for online learning; and (4) carrying out online training and updating for the initial ...

Process for azine

Hub motor and drum brake integrated device

The invention discloses a hub motor and drum brake integrated device, which comprises a hub motor and a drum brake structure, and is characterized in that the hub motor comprises a rotor, a stator and a shell, the stator is fixed in the shell, and the rotor is arranged at the inner side of the stator; the drum type brake structure comprises a brake drum, a friction plate, brake shoes and a brake motor, the brake drum rotates along with the rotor, the two brake shoes are symmetrically arranged on the inner side of the brake drum, the friction plate is fixed to the outer wall faces of the brake shoes, and the brake motor is arranged between the two brake shoes. The brake motor drives the two brake shoes to move back to back and enables the friction plates to abut against the inner wall of the brake drum, and braking of the brake drum is achieved. According to the integrated braking device, an integrated structure with two braking modes is adopted, the advantages of high response speed and ...

Novel split type disc friction brake

A first brake block and a second brake block are symmetrically arranged, the first brake block is fixed to a nut, the second brake block is fixed to a brake caliper, the brake caliper is connected to a fixing base in a sliding mode, a guide rod horizontally penetrates through the brake caliper and the nut, and the guide rod is connected to the fixing base in a sliding mode. Two transmission threaded surfaces with opposite spiral directions are arranged on the screw rod and are respectively in threaded connection with the brake caliper and the nut, and the motor drives the screw rod to rotate, so that the first brake block and the second brake block simultaneously move in the same direction or in the opposite directions; the brake disc is of a split structure and comprises an outer disc body and an inner disc body. Pure mechanical braking is adopted, a traditional hydraulic or pneumatic pipeline is omitted, a sodium filling cavity is formed in the brake disc, metal sodium has the characteristics ...

Titanium dioxide acid wastewater treatment and titanium gypsum comprehensive utilization method

The invention discloses a titanium dioxide acid wastewater treatment and titanium gypsum comprehensive utilization method, the wastewater treatment comprises the following steps: adding acid wastewater into limestone slurry, and carrying out first-step neutralization reaction to generate primary precipitated white gypsum and supernate; then carrying out acid leaching to remove iron, and washing, dehydrating, drying and calcining the precipitate to obtain gypsum powder; adding limestone slurry into the supernate, continuously stirring, carrying out a second-step neutralization reaction to generate a secondary precipitate and supernate, and dehydrating the secondary precipitate to serve as a raw material for recovering vanadium and titanium; and adding the alkaline residue slurry into the supernate generated in the second-step neutralization reaction to carry out third-step neutralization reaction to generate third-time precipitate, and dehydrating the third-time precipitate to be used as ...

Real-time bus passenger flow volume statistical method

The invention provides a real-time bus passenger flow volume statistical method based on a target prior distribution, which adopts the methods of target detection, target tracking, target behavioral analysis and the like, and belongs to the technical field of pattern recognition. The method concretely comprises the following steps: utilizing the target prior grayscale to do statistics of the back-projection operation of a histogram on an input video image, and then performing the treatment of difference, binaryzation, wave filtering, connected domain marking and the like on a back-projection image to realize the target detection; adopting the grayscale to interrelate, match and track and a mean translation algorithm to search and track so as to realize the accurate positioning of the target; and finally analyzing the motion trail of the target to judge the get on or off of passengers, and realizing the passenger flow volume statistics. The invention can provide fine-grained reliable basis ...

Fully-autonomous on-line study method based on random fern classifier

The invention discloses a fully-autonomous on-line study method based on a random fern classifier. With the method, on-line study targeted for a classifier of the target class can be carried out by selecting a target in a video frame for only once. The method comprises the following steps: carrying out affine transformation on the selected target to obtain an initial positive sample set, and extracting a small number of negative sample sets in a non- target area of the video to train an initial random fern classifier; then, carrying out target detection in the video frame by using the classifier; in the detection process, using a nearest neighbor classifier to collect on-line study new samples and automatically judging the class of the samples; and finally, applying the new samples to on-line training of the random fern classifier, updating random fern posterior probability, improving the precision of target detection of the classifier gradually, and realizing fully-autonomous on-line study ...

Disc type integrated braking device of electric automobile

According to the technical scheme, the disc type integrated braking device of the electric automobile is characterized in that the disc type integrated braking device comprises a hub motor rotor and a hub motor stator which are arranged on the inner side of a tire and the inner side of a hub, the hub motor stator is arranged on the inner side of the hub motor rotor, and a half shaft is fixedly connected to the middle position of the tire and the middle position of the hub; the inner side of the hub motor stator is rotationally connected with the half shaft through a first bearing, and a rotor magnet yoke, a permanent magnet and a stator winding are arranged between the hub motor rotor and the hub motor stator. A permanent magnet, a power generation stator coil mounting bracket and a power generation stator coil are also arranged between the hub motor rotor and the hub motor stator; functions and advantages of an electromagnetic braking system and a drive-by-wire friction braking system ...

Target detection method

The invention discloses a target detection method based on weak linear regression trees with incremental dimensions, which comprises the following steps: (1) preparing a training sample set (xi, yi), wherein i is equal to 1,..., N, xi expresses the characteristic values combination of training samples, yi expresses the types of samples, N expresses the number of the training samples, and N is a natural number; (2) initializing the weight of each training sample to be as shown in the specification, wherein t is a natural number and is equal to 1 in the process of initialization. (3) circularly computing the sample set for T times to acquire a linear regression tree every time to be used as a weak classifier, and then, synthesizing the T numbered weak classifiers acquired by circularly computing for T times into a strong classifier, wherein T is a natural number; and (4) classifying each region in digital images by using the strong classifier to judge whether the region is a target region ...

Kit and method for rapidly detecting vibrio tower, vibrio mediterraneus and vibrio harveyi

The invention provides a kit and a method for rapidly detecting vibrio tower, vibrio mediterraneus and vibrio harveyi. The method comprises the following steps: firstly, designing a series of primer probe combinations aiming at three pathogenic bacteria of vibrio tower, vibrio mediterraneus and vibrio harveyi, and carrying out multiple rounds of screening to find a triple amplification primer probe group with the highest detection sensitivity; meanwhile, an amplification system is optimized, and oligopeptide and DMSO are added into the A Buffer, so that the amplification efficiency is remarkably improved, the detection sensitivity and specificity are improved, 0.5 copies/mu L of samples can be detected at least, the specificity is good, and the detection time is only 10-20 minutes. The method is of great significance to rapid detection of tower-type vibrio, vibrio mediterraneus and vibrio harveyi, cutting off of propagation paths and reduction of the risk of industrial aquaculture of Babylonia ...

Pedestrian re-identification method and device based on mixed attention decoupling re-identification network

The invention provides a pedestrian re-identification method and device based on a mixed attention decoupling re-identification network, so as to enhance the field-invariant pedestrian feature discrimination capability, thereby forming a reliable category boundary and learning intra-class semantic diversity. The design of a mixed attention module in the method is to strengthen domain-invariant feature expression in an attention weight decoupling mode from the angles of space and channels, which forces the network to automatically utilize image regions and attribute clues beneficial to cross-domain re-identification. Besides, based on enhanced domain invariant feature expression, a multi-difficult sample memory learning strategy is provided to improve intra-class diversity of target domain samples. According to the method, the feature learning process is optimized by updating the reliable sample memory bank and the multiple difficult sample memory banks, significant intra-class semantic ...

Method for counting passenger flow of buses in real time

The invention provides a method for counting passenger flow of buses in real time, which adopts means of machine learning, multi-target detection and tracking, target behavior analysis and the like, and belongs to the technical field of pattern recognition. The method comprises the following concrete steps: detecting passenger targets by analyzing shape information and texture information of passenger heads in video images, wherein a column diagram in a gradient direction is used for representing the shape information, and a column diagram in a local binary mode is used for representing the texture information; then, accurately positioning passengers by adopting the target tracking policy combining gray level cross-correlation matching tracking and mean translation algorithm searching tracking; and finally, judging the behavior characteristics of the passengers by analyzing the moving tracks of the passengers, thereby accurately counting the passenger flow of buses. The practice showsthat ...

Real-time bus passenger flow volume statistical method

The invention provides a real-time bus passenger flow volume statistical method based on a target prior distribution, which adopts the methods of target detection, target tracking, target behavioral analysis and the like, and belongs to the technical field of pattern recognition. The method concretely comprises the following steps: utilizing the target prior grayscale to do statistics of the back-projection operation of a histogram on an input video image, and then performing the treatment of difference, binaryzation, wave filtering, connected domain marking and the like on a back-projection image to realize the target detection; adopting the grayscale to interrelate, match and track and a mean translation algorithm to search and track so as to realize the accurate positioning of the target; and finally analyzing the motion trail of the target to judge the get on or off of passengers, and realizing the passenger flow volume statistics. The invention can provide fine-grained reliable basis ...

Novel brake device of hub motor automobile

According to the technical scheme, the novel brake device of the hub motor automobile is characterized by comprising a hub motor outer rotor arranged on the inner side of a tire and a first bearing connected with the inner side of a hub, the hub motor outer rotor is rotationally connected with the first bearing, and the automobile hub is fixedly connected with a half shaft through the first bearing; an automobile tire and a hub rotate along with the half shaft, a second friction brake disc is arranged on the outer side of the first bearing, the second friction brake disc is rotationally arranged in the hub motor outer rotor, a rotor magnet yoke is arranged on the inner surface of the hub motor outer rotor, and a hub motor permanent magnet is arranged on the inner side of the rotor magnet yoke; the hub motor, the permanent magnet brake and the friction brake are integrated in the hub, the advantages of the hub motor, the permanent magnet brake and the friction brake are fully utilized, and ...

Composite brake device of hub motor vehicle

The invention discloses a wheel hub motor vehicle composite brake device which comprises a wheel hub motor, a friction brake disc, an electromagnetic brake disc, a brake caliper body, a power generation stator coil, a permanent magnet and an electromagnetic assembly. The requirements of different braking modes for the material and performance of the brake disc are met, meanwhile, the three braking modes of hub motor regenerative braking, electromagnetic braking and friction braking are integrated to meet different braking requirements, on the premise that braking safety is guaranteed, hub motor regenerative braking is preferentially adopted, and the braking efficiency is improved. And when the regenerative braking torque cannot meet the braking requirement, electromagnetic braking or friction braking participates in braking, response is conducted within the shortest time, and the vehicle is decelerated or stopped within the shortest braking distance. And meanwhile, electric energy required ...