At present, beer production enterprises pay more and more attention to product quality, and more and more companies use various automatic testing equipments in filling production lines in order to increase the reliability of detection. The traditional bottle filling inspection method is artificial light detection. However, in modern high-speed filling lines, due to visual fatigue and other reasons, reliability cannot be ensured only by traditional artificial light detection, and two important parts of the beer bottle, the bottle mouth and the bottom of the bottle, cannot be detected, resulting in partial leakage. Bottles and bottles at the bottom of the bottle flow into the market, causing damage to the corporate image.


1 Online detection of basic conditions


Currently, on-line detection of empty bottles based on image processing technology abroad has been relatively mature and there are a large number of products on the market, such as: MIHO system, HEUFT system in Germany; OMNIVISIONI system of INDUSTRIALDYNAMICS COMPANY, LTD in the United States.


At present, most of the beer manufacturers in China are developing in the direction of production lines of more than 40,000 bottles per hour. It will be extremely difficult to rely on manual inspections at this speed, and they urgently need a large number of empty bottle inspection equipment. The price of foreign products is too high, and maintenance is difficult. The threshold is too high for most small and medium-sized breweries.


2 Online detection of target determination


On the basis of reference to the models of foreign companies, analyze the needs and technical strengths of current Chinese SMEs and determine the objectives of online testing:


2.1 Online detection module:
The bottle mouth and detection bottom are used as basic modules. Additional modules such as inner and outer wall detection can be reserved.


2.2 Online detection speed:
The bottle neck is 64mm with 42000 bottles/h.


2.3 Bottle bottom detection resolution and reject rate:
No mark and no wear bottle: When the size is 2.5mm×2.5mm, the rejection rate is 98%; when the size is 3.0mm×3.0mm, the rejection rate is 99%. Signage, wear, dark, color bottle: When the size is 3.5mm×3.5mm, the rejection rate is 98%; when the size is 4.0mm×4.0mm, the rejection rate is 99%. Can handle mixed or dark bottles.


3 Determination of the online testing program

3.1 On-line detection scheme Option 1: Direct purchase of a foreign machine vision system, based on the system for secondary development. Although the development speed of this program is fast, the machine vision system is expensive, and the software registration fee must be paid for each subsequent production. The ready-to-use vision system that can meet the speed requirement also hardly exists. This solution is not desirable.


Option 2: A hierarchical control system is adopted, that is, an industrial control machine completes the upper-level control and management work, and a lower-level system is equipped with a DSP chip-based system to complete image acquisition and image processing.
Due to the long development cycle, this solution is not conducive to occupying the market, and the plan is based on the fact that the algorithm has been fully matured. Therefore, the plan is postponed for future improvement.


Option 3: Use the image card industrial control computer and self-developed control software. In consideration of the processing speed, a multi-system industrial control computer is required. This program is economically, technologically and practically suitable for SMEs, and has already been successfully studied by some beer production companies. Therefore, the plan is adopted.

3.2 Online Inspection Workflow

The empty bottle to be inspected is sent by the conveyor belt. When it reaches the position of the mixed bottle and the color difference compensation light source, the light source flashes. The mixed bottle and the color difference sensor detect the transmitted light and send the detection result to the color difference compensation circuit so as to detect the light source at the bottom of the bottle. Brightness compensation. When the empty bottle reaches the bottle bottom detection position, under the action of the color difference compensation circuit and the camera light source synchronization circuit, the light source is flashed and the shutter is opened, and the camera is imaged by the imaging system. The resulting image is sent to the bottom computer for processing by an image capture card. If there is dirt after processing, a reject signal is sent through 32 I/O cards. The signal is amplified and the control reject mechanism pushes the bottle away from the conveyor belt. When the empty bottle enters the bottle mouth treatment position, the bottle mouth treatment is performed, and the treatment process is basically similar to the bottle bottom.


The bottle bottom handling master control management computer and the bottle handling computer communicate with each other through the serial port to ensure system coordination. The two systems share the same touch screen for display and operation, and the switching between the two systems is accomplished through a touch screen switching circuit.


3.3 Bottle mouth (bottom treatment module) processing module working principle

3.3.1 Bottle mouth detection We use a special imaging device to make the flash light shine on the bottle sealing surface and return to the camera. An intact plane reflecting light should be a complete and smooth ring. If it is damaged, the ring breaks (through breakage) or the ring appears concave (breakage of the inner edge) or the ring appears convex (breakage of the outer edge), The user can define the size of the fracture, indentation, and outburst by setting three limit values.
The system determines whether to reject by comparing the size of the actual fracture, concave, and convex with the set value. In order to avoid misjudgment caused by the bottle tilting, the module can automatically track the position of the bottle via a certain method.

3.3.2 Bottom detection Firstly, the grayscale image obtained by the image acquisition card is binarized, and the size of each independent stain is calculated by a statistical method. The value of the stain size is compared with the value of the user-defined allowable stain, and the decision is made. Is it removed?

4. Conclusions The use of image cards and multi-system industrial control computers has solved the problem of online detection of defective beer bottles, and the price is relatively low, which can meet the needs of most small and medium brewers and achieve better economic benefits.

20 Pairs Lashes

20 Pairs Lashes,20 Pairs Fake Lashes,Strips Fake Eyelashes,Natural Classic Lashes

Zhengzhou Cuka Electronic Commerce Co., Ltd. , https://www.cukalashes.com