Machine vision-based instrument panel assembly intelligent integrated test system
EQ140-II automotive instrument panel assembly is an automotive company in China's production of instrumentation products, the instrument panel is installed with the speedometer, water temperature, gasoline gauge, ammeter, signal alarm lamps, etc., the production of its production batch is large, the factory needs to be carried out before the final inspection of a quality. Inspection items include: testing the speedometer and other five gauges pointer indication error; testing the 24 signal alarm lamps and a number of lighting 9 lamps are damaged or missing. Generally, the manual visual inspection method is used, which has large error and poor reliability, and cannot meet the needs of automated production. Intelligent integrated test system based on machine vision changes this status quo, realizes intelligent, fully automatic, high precision and fast quality inspection of instrument panel assembly, overcomes various errors caused by manual inspection, and greatly improves the inspection efficiency. The whole system is divided into four parts: an integrated multi-channel standard signal source that provides an analog signal source for the instrument panel, a dual-coordinate CNC system with image information feedback positioning, a camera image acquisition system and a master-slave parallel processing system. Metal plate such as large power transformer coil flat line radio mask hazel skin surface quality have high requirements, but the original use of manual visual or with a percentage meter plus control needle detection method is not only susceptible to the influence of subjective factors, and may draw the surface to be measured to bring new scratches. Sheet metal surface automatic flaw detection system using machine vision technology for automatic inspection of metal surface defects, high-speed, accurate detection in the production process, and at the same time, due to the use of non-contact angle measurement, to avoid the possibility of generating new scratches. The working principle diagram is shown in Figure 8-6; in this system, a laser is used as the light source, the stray light around the laser beam is filtered out by a pinhole filter, and the beam expander and collimator mirrors make the laser beam into parallel light and illuminate the surface of the metal plate being inspected uniformly at an incidence angle of 45 degrees. The plate is placed on an inspection table. Inspection table can be moved in the X, Y, Z three directions, the camera adopts TCD142D type 2048 lines Chen CCD, the lens is used in ordinary camera lenses, CCD interface circuit using a microcontroller system. The host PC is mainly to complete the image pre-processing and defects in the classification or depth of the scratch operation, and can be detected defects or scratches in the image display monitor.CCD interface circuit and PC through the RS-232 port for bidirectional communication, combined with the asynchronous A/D conversion, constituting a man-machine interactive data acquisition and processing.
The system mainly utilizes the self-scanning characteristics of the line array CCD combined with the X-direction movement of the inspected steel plate to obtain three-dimensional image information of the metal plate surface. British ROVER Automotive 800 series car body contour dimensional accuracy of 100% online inspection, is the machine vision system used in industrial inspection of a more typical example, the system consists of 62 measurement units, each measurement unit includes a laser and a CCD camera to detect the body shell on the 288 measurement points. The car body is placed under the measuring frame, which is calibrated to the exact position of the body by software.
Special attention is paid to the calibration of the measurement units, which affects the accuracy of the inspection. Each laser/camera unit is calibrated offline. There is also a calibration unit that has been calibrated offline with a CMM, allowing the camera top to be calibrated online.
The inspection system inspects three body types at a rate of one body every 40 seconds. The system compares the inspection results to qualified dimensions summarized by a person, from a CAD model, with a measurement accuracy of ±0.1mm. ROVER's quality inspectors use the system to discern the dimensional consistency of critical parts, such as the overall body appearance, doors, and glass window openings. The system has proven to be a success and will be used for bodywork inspection on other ROVER system cars. Voice from the end user - costing should be viewed in a developmental perspective .
Machine vision technology has not been widely used mainly because of the high cost of the product. Has not yet formed a widely used environment. But from the perspective of development. Long-term accumulation of manual inspection costs will be equal to the machine vision equipment. In the high-speed production line to match the machine vision equipment will be a more obvious development trend. Along with the gradual upgrading of pharmaceutical enterprises and gradually improve. The demand for machine vision technology will become more and more common. In addition. From the perspective of social responsibility. More and more enterprises in the pharmaceutical industry will "guarantee the safety of patients' lives" and "the fulfillment of corporate responsibility" as the supreme mission. This is driving the use of machine vision technology in the industry.
Voices from suppliers - responsibility, technological development, policy requirements, 'none less'.
Many domestic pharmaceutical companies in the production of drugs want to keep costs as low as possible, so that. The use of some of the corresponding testing equipment is limited to some extent. Pharmaceutical companies to reduce the cost of production considerations is the key factor restricting the wide application of machine vision technology in the pharmaceutical industry.
And on the other hand. Pharmaceutical industry characteristics also determine the urgent need for machine vision technology. If the enterprise has very high requirements for itself. The quality of the product and social responsibility has a very responsible attitude. Then they will not hesitate to apply machine vision technology. Because they know where to do what. Which money to spend.
From a technical point of view. There are many inspection technologies for pharmaceutical packaging. Machine vision technology is not the only one. But for some of the more complex or specific inspection programs. Machine vision technology can indeed give the safest and clearest detection. Pharmaceutical companies in the inspection equipment can achieve the same quality under the premise. They can choose the equipment they think is more suitable. But if they give up quality for some reason. Choose the detection effect is not the best or can not fully meet the detection requirements of the equipment. This is not worth the loss.
Analysis of the advantages of machine vision technology compared with other detection technologies
Voice from the end-user - manual inspection, photoelectric detection can no longer meet the detection needs of enterprises
After the reporter interviewed a number of pharmaceutical manufacturers were informed. Manual inspection in China's pharmaceutical companies are still widely used. Nu Skin Healthcare Pharmaceuticals China Huzhou production base of the production manager Mr. Lin Jie capsule dosage forms as an example to explain the working process of manual inspection: usually for the rotating capsule through the conveyor belt driven. The capsules are usually rotated on a conveyor belt, passed by an operator and inspected manually by visual inspection. There is no doubt that this traditional way cannot avoid some objective risks. For example, human eye fatigue, errors, high speed production process of energy concentration, the inability to quantitative assessment of the detection effect, speed limitations and so on. And these unfavorable factors will undoubtedly become a bottleneck in the future development of pharmaceutical companies. Packaging itself involves a problem of identification. Through some reliable machine identification technology will achieve more stable results than manual identification. In this case. Machine vision technology came into being. And will win a broader market. At the same time. In the high-tech concentration of the number of particles in the link. Photoelectric inspection equipment for imaging, counting process is not as authoritative and accurate as machine vision technology. --More comprehensive, reliable machine vision technology in the future will replace other related detection technology
In addition to machine vision technology. Manual inspection, photoelectric technology, weighing technology and other inspection technologies are also used by pharmaceutical companies. But compared with other detection. Machine vision technology in the detection accuracy and operational simplicity are better. For example, photoelectric technology is more complex in terms of commissioning and setting. The slightest change in the details of the packaging line will bring inaccurate detection results. For example: displacement changes, vibration. Because the pharmaceutical packaging line is a high-speed packaging line. Complexity. So these situations often occur. Pharmaceutical companies to find the problem and then debug the testing equipment settings will be more difficult. And if the pharmaceutical company does not find these problems in the test. Then there may be packaging problems in the product put on the market to sell. The consequences are unimaginable.
In addition, the photoelectric detection of the detection program is very single. Can only detect a project. This requires the installation of multiple photoelectric detection equipment on the packaging line. For example: in the packaging detection of bottled medicine. Bottle placement, liquid level, labeling will require three photoelectric detection equipment to complete. But if the use of machine vision equipment can be a device to complete a test. And machine vision equipment in the application. As long as in the line of sight can be detected. Then some of the production line shift, vibration and other issues will not affect the final detection results.
Special mention. Machine vision equipment is also very suitable for pharmaceutical companies often change the packaging line packaging products. For example, a pharmaceutical company's blister packaging line may be producing a drug for stomach problems one day. The next day it may be producing a drug for heart disease. The two drugs will be inspected differently. The machine vision equipment can store more than a dozen products in the software. When you change products, you just need to call up the corresponding content. And do not need to re-adjust the parameters of the test. Waste a lot of time.