The invention provides a natural image target material visual feature mapping method based on a generative adversarial network. A deeply unsupervised learning way is used for learning the unlabeled natural image target material visual feature to obtain the high-order expression of image target material visual feature space, a mapping network about the material visual feature space between a sourcedomain image and a target domain image is learnt and established, the material visual feature of the source domain image is mapped to the material visual feature of the target domain, so that the target domain image has the material visual feature information of the source domain image, and finally, an image of which the material visual feature is mapped is obtained. The method learns from the unlabeled natural image to obtain the material visual feature information, the task target of which the material visual feature is mapped among different images is carried out, a corresponding solutionis put forward by aiming at the task target, a good result is obtained, and the method has important theoretical significance and practical values.

Natural image target material visual feature mapping method based on generative adversarial network

The invention discloses a material recognition method of the MIMO radar based on machine learning. The method comprises steps of firstly, using the MIMO radar to collects echo signals of sample materials; carrying out feature extraction of the echoes to obtain feature parameters capable of differentiating different materials; using the machine learning algorithm to determine mapping relations between different materials and feature parameters; and finally, according to the mapping relations, determining which kind of materials that an object detected by the MIMO radar belongs to. The method is highly effective, feasible, high in performance and capable of recognizing materials precisely.

Material recognition method of MIMO radar based on machine learning

The present invention provides an object identification method based on room acoustics channel disturbance analysis. The method is completed through two phases. The first phase is the establishing of a sample database, and the second phase is a realistic identification phase. In the room application environment scene, different types of objects for identification in future are selected, acoustics channels are measured when the objects are located at different positions in a room, the measurement signals are subjected to feature extraction to summarize the features of different objects, and a result is taken as a sample feature library; and in the realistic identification process, the acoustics channel is measured again when there is a certain object, and a measurement result is subjected to feature extraction according to the feature extraction method and is subjected to cooperation processing with data in the sample database so as to finally identify the objects in the room. The object identification method based on the room acoustics channel disturbance analysis has the advantages of non-visual identification and adoption of few hardware devices compared to widely applied video identification and WIFI identification. The object identification method based on the room acoustics channel disturbance analysis fully utilizes the acoustics information of the room channels, is simple to calculate, and high in location efficiency.

Object identification method based on room acoustics channel disturbance analysis

The invention discloses a mistake-proof detection method for a structural member assembly in manufacturing industries of automobiles, machines or home appliances. The mistake-proof detection method comprises the following steps: I, defining a structural member assembly to be detected; II, setting a detection mold, and arranging the detection mold in detection equipment, wherein the detection mold is matched with the size of the structural member assembly to be detected; III, arranging a plurality of photoelectric switches in the detection equipment; IV, arranging a plurality of RGB digital optic fiber sensors in the detection equipment; V, arranging a plurality of digital laser sensors in the detection equipment; VI, calibrating the RGB digital optic fiber sensors and the digital laser sensors; VII, placing the structural member assembly to be detected in the detection mold, starting the detection equipment and an industrial control computer, analyzing and processing received data by the industrial control computer, and outputting a detection result to a display for displaying. The mistake-proof detection method can be widely applied to automated test systems in the fields of the automobiles, the motorcycles, the machines, the home appliances and the like.

Mistake-proof detection method for structural member assembly in manufacturing industries of automobiles, machines or home appliances

The invention discloses a material acquisition method and system based on a neural network. The method comprises a training stage and a collection stage. In the training stage, the parameters of the acquisition equipment are firstly obtained, and the synthetic material data which simulates the actual camera acquisition results are generated as the training data; then the neural network is trainedaccording to the synthesized training data. In the collection stage, the target object is illuminated according to the illumination pattern, and a group of photos are obtained, and then, the pixels and channels of the photos are traversed in turn to form a vector and sent into the restoration network to obtain the material information. The invention greatly reduces the time required for material quality collection and has higher collection quality. The acquisition condition is near-field light, and the acquisition object is anisotropic material, which is the most complex case in the field, andthe method of the invention only needs a few photographs. Any collection device using the illumination pattern idea can utilize the method of the invention to train the network pertinently and improve the collection efficiency of the high-quality material information.

A material acquisition method and system based on neural network

The invention provides an object material identification method based on multi-sensor information fusion. The method comprises steps: firstly, at least three kinds of sensors are provided; then, the sensors are used for transmitting source signals to a to-be-identified object, feature signals transmitted by the to-be-identified object under irradiation of the source signals are acquired respectively, or the sensors are directly used for acquiring the feature signals of the to-be-identified object, and object material information provided by each feature signal is extracted; and finally, the information is analyzed comprehensively to determine the material of the object. According to the object material identification method of the invention, data complementation and redundancy of different sensors are used, information is acquired from each independent piece of measurement space, the target object material can be identified through a fusion technology, the detection accuracy is high, and reliable data are provided for object material identification.

Object material identification method based on multi-sensor information fusion

The invention provides an image collection system for assisting airplane inspection and remote fault diagnosis. The image collection system comprises an image collection device which is used for collecting images of an airplane, wherein the image collection device comprises an image collection module fixed on a mobile device, and an image generation sending module connected with the image collection module; a remote processing decision-making center which is used for storing the images and realizing airplane inspection and fault diagnosis through a special expert system; and a data communication network which is connected with the image collection device and the remote processing decision-making center and is used for sending the images collected by the image collection device to the remote processing decision-making center. Through adoption of the image collection system for assisting airplane inspection and remote fault diagnosis provided by the invention, the inconvenience of inspecting the airplane through adoption of an on-site manual mode in the prior art is solved, and the method for centrally processing inspection and fault diagnosis and establishing visual inspection records is provided.

Zhou Han

Papers

Object material identification method based on multi-sensor information fusion

Image collection system for assisting airplane inspection and remote fault diagnosis