I'm taking over the development of an intelligent detection system for a telecommuting area of a smart factory.

 Success Stories    |      2021-06-10

1、           Arithmetic for In-depth Learning

Integrated use of identification, testing, tracking and video structured analysis to translate video content into text messages that can be understood by computers and users; thus, integrated perception of scenes, people and objects; behaviour detection and prediction can also be achieved through video analysis, with the use of deep learning techniques to make attitude estimates of targets, match database behaviour templates, complete identification and generate results。

The main steps adopted are as follows:

(1)       Large-scale model training

(2)       Auto-identification of models

(3)       Tasks and dispatches

 q1


2、Image and video recognition analysis

Video images are collected using cameras to calculate personnel behaviour patterns, identify personnel behaviour, identify staff attendance, normal walking behaviour, monitor area concentrations, crowding behaviour, face recognition and voice recognition, and identify the scene intelligence in and out of the commuter area as a manageable, real-time process data。

q2

Image and video recognition analysis is centred on video, combining artificial intelligence techniques, and the expansion of multi-dimensional sensing tools to identify individuals and objects entering the perimeter, with the following main identification effects::

(1)       Walking the tracks

(2)       Walking in the direction.

(3)       Arms wide.

(4)       Head test

(5)       Face recognition.

 

 

3、       Large data management and analysis

    Using large-scale parallel processing frameworks, super-calculation capabilities, further integration or integration of multiple complex isomer data sets, generating data collection, collation, cleansing, conversion from different data sets into a new data set, providing a unified data view for subsequent query and analytical processing, refining and establishing a calculation model from a diverse array of big data problems and needs based on different data features and computational features。


    In the application of big data analysis, visualization and visual analysis helps people to explore and understand complex data through interactive visual expressions, helps users to cross-check a large number of data and helps users to obtain new and faster discoveries from complex data.。


    For many actions, the very difference is very large, for example, when people walk at different times, they vary in speed or length. There may be a great similarity between different movements.

The context is dynamic or light, video shakes, etc., which influences the outcome.。

 

Collect three types of routine behavioural characteristics, as follows::

(1)10All kinds of routine movements: walking, running, jumping forward, testing, bending, waving, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, jumping, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing, dancing10Individual presentations, background fixed and outlook profile included in database, perspective fixed。

 

(2)6Action: Go, jump, run, box, wave, clapping, by25Individual execution, four scenarios, each.599The video, except for the closeness of the camera and the slight motion of the camera, is relatively static.。

 

(3)Organisation10Sports-type movements: diving, golfing, kicking, lifting, riding, running, skateboarding, swinging, side-stating, walking, covering a wide range of scenes and perspectives。

0