About the detection of refractory bricks

Refractory bricks are the basic construction materials in the high-temperature smelting industry, and the quality control of Refractory bricks is very important for manufacturers. The traditional quality inspection is to manually inspect each batch of products, which is labor-intensive and prone to missed inspections. With the development of machine vision and image processing technology, items such as fabrics, steel belts, masks, etc. have achieved automatic detection based on machine vision, with high efficiency and low cost. Below is information about the quality inspection of refractory bricks, dimensional measurement algorithms, and surface defect detection.

The following is a list of contents:

·Research background and significance of refractory bricks quality inspection

·Dimensional measurement algorithm and surface defect detection of refractory bricks

Research background and significance of refractory bricks quality inspection

Refractory bricks, as the most common refractory material, have regular shapes and standard sizes and are fired from refractory bricks or other refractory raw materials. Not only the refractoriness is above 1580 degrees, but it can also withstand various physical and chemical changes and mechanical effects. It is widely used in high-temperature industrial kilns and structural materials of various thermal equipment. It can improve the performance and efficiency of refining. It is an industrial production process. the most important line of security. The production process of refractory bricks is very complex, and the quality of the products must be strictly controlled at each stage. Improper molding and firing may cause defects in refractory bricks. Refractory bricks are closely related to high-temperature industrial production. Its quality and life span will affect the economic benefits and production safety of the entire industry. Recently, there are still problems such as low production efficiency, low industrial concentration, and low level of intelligence. The company will continue to use new technologies to develop quality inspection methods, improve the level of intelligence and automation, and promote the digital upgrade of the industry.

Quality inspection is an indispensable and important step in the production process of refractory bricks. From the stirring of raw materials to form, there will always be some unqualified waste products in the size and appearance of refractory bricks after firing. To ensure the quality of production, the semi-finished and fired products of refractory bricks must be screened for size and surface defects according to national standards. The picking workers use measuring tools such as rulers to manually measure the size of the refractory bricks, and at the same time judge by experience, whether there are defects such as missing corners, flash edges, and cracks. Workers need to stay on the production line for a long time, the work intensity is high and the content is mechanically repeated. The detection is completely based on experience, there is no unified evaluation standard, there is a certain subjective impact, and it is easy to cause false detection and missed detection, so the detection efficiency is low. Due to the limitation of manpower and cost, the quality inspection of refractory bricks usually adopts the method of random inspection, which cannot be fully inspected. Once a defect caused by equipment failure occurs, it will cause huge economic losses. At the same time, if the waste products flow into the market, it is a big safety hazard for the high-temperature industrial industry. To effectively reduce the defective rate of refractory bricks, the company must first start with the detection method, improve the efficiency of detection, and timely detect equipment failures on the production line, to better control the product quality.

Dimensional measurement algorithm and surface defect detection of refractory bricks

According to the characteristics of refractory brick defect images, our company has designed a size measurement algorithm based on image processing and a surface defect detection based on deep learning. In the size measurement of refractory bricks, first, use the preprocessing method to extract the contour, and then use the fine contours function of OpenCV to find the contour in the image, the min Arca Rect function fits the shape of the refractory bricks and calculates the refractory bricks according to the coordinate points of the external matrix. Width and length in pixels. In surface defect detection, a defect detection model for refractory bricks is designed based on the inception structure block. The model has 13 layers, which is more streamlined than the original GoogLeNet. Finally, use the refractory bricks surface defect data set to train the refractory bricks to defect detection model and GoogLeNet respectively. The refractory bricks model proposed by our company can achieve an accuracy rate of 94%, and there is no over-fitting phenomenon, which is better than GoogLeNet.

The research on defect detection of refractory bricks has great significance for the whole industry. The company's research on defect detection of refractory bricks is not limited to surface defects, and will also research internal defect detection in the future. After reading the above content, if you are interested in refractory bricks, you can check the website https://www.yufengrefractory.com/ of our company, we look forward to your arrival.