To obtain perfect joins, inside of the solder ring copper fitting and the outside of the copper pipe are cleaned using coarse steel wool, flux paste is applied, the pipe is then inserted into the fitting and heat applied from a portable propane torch until a ring of solder shows at the edges of the fitting. Sizes we have available are from 8mm to 54mm For EN1254-1 and 1/2" to 2" for ASME B16.22 ,Solder Ring Copper Fittings,presolder copper fittings,yorkshire copper fittings, pegler copper fittings , connex copper fittings solder ring copper fittings,presolder copper fittings,yorkshire copper fittings,pegler copper fittings,connex copper fittings Taizhou Runde Company , https://www.smartfittings.com
Get rid of face reliance on NEC to launch new recognition technology [full text]
Technology News Recently, NEC announced that it has completed the development of a "Person Re-identification Technology" (Person Re-identification Technology), claiming that even if the camera cannot capture the complete image of the person (including the back and side faces), it can also be completed Matching of character images.
The technology will analyze the image of the person's photo, including the overall appearance, clothes, body shape, etc., to determine whether it matches the other images. Through NEC's facial recognition technology, image analysis, and deep learning technology, the establishment of not only depends on Face image recognition technology.
In actual use cases, facial recognition technology alone is not sufficient to match people whose faces are deliberately obscured. Traditionally, to use facial recognition technology to match a target person, you need to collect from the images taken by multiple cameras, but when multiple people or other objects appear in a scene, it is difficult for the camera to match the target person. Image data.
But this new technology supports the ability to recognize targets in situations where many people or visual obstacles prevent people from acquiring the faces of target people. According to reports, the potential applications of the technology include large public scenes and the search for missing children. According to Masayuki Mizuno, general manager of the NEC Biometrics Research Institute, combined with the strength of NEC in biometrics, the technology will solve more difficult problems for users in urban security solutions.
It is understood that the main feature of this technology is that when the camera cannot capture complete face information, the NEC-based deep learning algorithm automatically selects other images for matching through the overlap between multiple people. This is suitable for crowded spaces or The complete image of the target person cannot be viewed in a shadowy environment. According to the data, at present, the technology has an accuracy rate of 90% in matching characters.
Both the back and side faces can be matched, which means that based on deep learning technology, people can match the target person from multiple angles, and in some camera blind spots, they can also better perform crimes and crimes. Deterrent effect.