With the development of artificial intelligence technology, machine vision, with its fast, accurate and efficient information collection and processing advantages, has gradually become one of the indispensable technologies for the modernization and upgrading of manufacturing enterprises, and continues to empower the automation and intelligence of related traditional labor-intensive industries Transformation.
Traditional labor-intensive industries have a high degree of dependence on labor, long supply chain cycles, harsh working environments, and high labor levels of workers. In the wave of intelligent manufacturing, digital intelligence upgrades are urgently needed.
Textile industry needs transformation, AI machine vision to report
As a typical labor-intensive industry, the textile industry is dominated by traditional manual inspection methods. However, under the influence of modern production methods, the requirements for production technology are gradually becoming higher, and more attention is paid to digital and intelligent inspection methods.
Manual inspection has obvious disadvantages
High labor intensity: The thread breakage detection in the traditional textile industry needs to rely on spinning workers walking back and forth in front of the spinning machine to detect yarn breakage in time, and the labor intensity of workers is relatively high;
Poor workshop environment: high temperature and high humidity in the production workshop, accompanied by dust and noise pollution, have a greater impact on the health of workers;
· Difficulty in detection: If multiple yarns are broken, it is difficult to manually judge which yarn needs to be connected first, resulting in waste of time and raw materials;
· Product quality is different: due to the influence of work experience, fatigue and other factors, it is impossible to accurately calculate the spinning cycle of each bobbin due to yarn breakage, and it is also difficult to accurately control the yarn quality and count the yarn output. The qualified rate of products is only 60%-70%;
Therefore, the textile industry urgently needs to find a more efficient detection method to control production costs and improve product quality.
In view of the current situation of the above industry, many textile enterprises urgently need to implement the actual production line for automation and intelligent quality inspection. Coupled with the country's strong support for intelligent manufacturing, a large number of AI machine vision companies have emerged. While adhering to big customers and big-scenario tracks, these AI machine vision companies have always insisted on the exploration of small-scenes, and provided machine vision products and solutions for multiple scenarios with continuous upgrading of technology. Deep Eye Technology is one of them. .
Qingzhe visual engine empowers practical application scenarios
As a provider focusing on one-stop AI vision solutions, Deep Eye Technology has been deeply involved in the machine vision industry since its establishment, focusing on industrial AI vision technology, continuously developing innovative machine vision applications and developing industry application scenarios.
In the scene of thread breakage detection in the textile industry, the textile machine often breaks suddenly during the spinning process. Once the thread breakage is not detected in time, it is easy to damage the equipment, which will directly cause the spinning system of the textile factory to fail. effective operation, resulting in economic losses. Through AI machine vision technology, the detection rate of disconnection can be greatly improved, the normal operation of the production line can be guaranteed, and the overall production capacity of the textile industry can be further upgraded.
For the project of thread breakage detection in the textile industry, Shenmo Technology went deep into the actual scene, excavated and analyzed the pain and difficulty of the project, including but not limited to the different number of coils and various thread colors corresponding to the carding machine; the structure of the on-site equipment is complex and cannot be achieved Excellent light environment and shooting angle; the staff frequently change the bobbins, and the detection equipment cannot fix the position; in order to ensure the quality control of the coil, it is difficult to reconnect within a short period of time after the wire is broken.
With the advantages of rich algorithm development experience, professional talents and 2000+ AI industry models, Shenmo Technology has developed a carding machine disconnection detection system for the textile industry disconnection detection scenario.
Among the many landing cases in the textile industry, an employee of an enterprise's intelligent workshop said: the software platform can analyze the captured images, detect yarn breakage in time and generate an alarm, and the staff can rush to the scene within one minute. And according to the analyzed image, locate the broken line position, and carry out timely processing. At the same time, originally one person needs to manage two looms, but after using this system, only one employee is needed for every 20 spinning machines on average, and the detection efficiency is increased by more than 10 times.
The carding machine disconnection detection system builds a business process through the Qingzhe visual engine, connects multiple cameras in series, and realizes scene multiplexing positioning and segmentation algorithms, and then integrates deep learning technology with machine vision to realize a large amount of learning about product features. Continuously improve the detection effect of disconnection identification in production.
Deep Eye Technology focuses on small but large potential demand scenarios. In the case of disconnection detection, based on the "traditional machine vision + deep learning" technology, it realizes in-depth analysis of the complex surface of textiles and solves the problem of artificial or traditional machine Disadvantages of insufficient visual recognition ability. In the future, Deep Eye Technology will continue to explore more small scenes and penetrate more industrial fields with AI vision solutions.