Testing of Pharmaceutical Tablet Strip

Testing of Pharmaceutical Tablet Strip CAREER EPISODE 1 1.1 INTRODUCTION I completed my Bachelors degree in electronics and communication engineering in the year 2011. I pursed my engineering from Vyas Institute of Engineering and Technology, Jodhpur, India. During my final semester of engineering I had to make a working model in the field of electronics. I came up with a project called Testing of Pharmaceutical Tablet strip. The project was performed in group of three. It was completed under the direction and support of Er. Yogesh Chaudhary. In my first career episode, I have described the work done by me in this project in detail. 1.2 BACKGROUND Requirement of medicines and its production is increasing rapidly. There are possible occurrences of breakage, cracks in the capsule or table while its production. And when these damaged tablets are consumed, it may cause irritation or side effects to other body parts. To inspect the production, it becomes very tough manually but the same task can be achieved with the help of image processing through automation of visual inspection. My project deals with the identification of the damaged products after its manufacturing. It involves series of task like segmentation, image processing, filtration, subtraction, pixel-calculation, de-noising, thresholding, and region based statics to identify the damaged and broken tablets. 1.3 The basic moto behind my project is to identify and filter out the damaged and defective capsule strips. Missing tablet, color, or shape/size difference between a set of tablets, or any crack or breakage in the tablets will be identified and marked as defect through my project. The idea was to reduce the manual work for identification of the damaged tablets as the job can be tedious. And to introduce a new procedure that would contribute in digital technology and for the welfare of the society. Identification of defective tablets is done with the help of image processing that uses various techniques of algorithm for processing all the digital images. 1.4 Image processing works on the technique that uses features like extraction, pattern recognition, edge detection and template matching. This process manipulates the data faster to achieve the desired result. For quality assurance of the products that are being manufactured majorly require automation of visual inspection. In this technique, a morphological operation such as opening operation is used for finding defects. The image is made fit for the further processing by performing image segmentation on the input image and then it is filtered to remove any noise. By inscribing rectangles subtraction is performed on the image with the help of morphological operations and it is also subtracted from original gray image which shows the broken tablets. For calculating the exact location of the broken tablet, pseudo coloring method is used and the pixel of the broken capsule is calculated. Correct size or color and any cracks are detected by performing corner detection and Harris algorithm technique on the tablets. Which is then followed by some pre-processing. After detecting the corners, they are compared with the template image and if there is any mismatch in the feature with the original template and the test image, the tablet gets rejected otherwise it is accept and given for use. 1.5 Me being the team leader, had to work out the flow and plan on how to proceed with the implementation process. I divided the project into small tasks and gave priority to each of them. Then they were assigned the duration in which that task had to be completed. My major task was designing the circuit diagram and block diagram on which the components could be decided. And the other crucial task was software coding. Regular meetings were kept with my team mates and the project guide to get any suggestions for any improvement in the project. Meetings were either through email or small presentations had to be made. 1.6 PERSONAL ENGINEERING ACTIVITY I started off with surveying the market so that the project I make would be useful. I read papers and articles relating the new upcoming technology and tried to do research on image processing methods and its technical specifications. This helped me to learn and build new technical skills. By applying my engineering knowledge and through technical reading I plotted a block diagram and decided upon the main hardware components that could be used for this project. It involved: Electrical components like Power supply of 230V, contractor 220-240V 1.47A, conveyor belt, step down transformer and DC motor. Electronics components such as a microcontroller, LCD, Regulator IC, relay. And a basic camera of 1.3MP/VGA With the help of the hardware components and block diagram I used a PCB designing tool for designing and structuring my circuit diagram. 1.7 After deciding and finalizing of the components, I wrote down the procedure in which the tracking of the missing tablet or damaged table would be done. The camera was used basically to get the complete image of the production. After getting the image, it is converted to monochrome image from RGB one. This results in the display of the tablet strips in black and white form, where the strips will be seen in grey color and the tablets would be in black. A predefined number is assigned and kept for the tablets in a strip to keep the track of it. If the tablets are found to be lower than the defined number after image processing takes place, the microcontroller will signal the second motor to discard the strip detected. This was the logical that was used for finding the tablet strip with missing tablet. 1.8 In the similar fashion a procedure was implied for detecting and finding cracks in the tablets. For that, firstly all other components are removed from the image and only the tablet is seen. This is done through filtration technique of image processing.   After this the image is converted to black and white from the original RGB image.   Because of such design and method, only cracks could be seen in the image if there are any. They would differ in color from the black tablet. They would appear to be white thin or thick lines. So, by applying such functioning technique, the presence of any crack in the tablet could be traced. 1.9 The working of my model is very simple. Above two logics are converted into a programming code through two software Bascom AVR and MATLAB 2013, and then the code is installed in the microcontroller ATmega16. Working with MATLAB software was not difficult and coding was not an issue because of having a subject of image processing in my curriculum. When the system is powered on with the help of power supply section, the conveyor belt starts to move. When the camera lens attached to the system senses the tablet strip that is passed through the conveyor belt, it would signal the microcontroller to stop the belt that runs with the help of motor 1. This is done for analyzing the tablet strip. ADC is used for converting the analog power signal from sensors to digital form for signaling the microcontroller. The image taken of the tablet strip is scanned and analyzed according[S1] to the functions as designed in MATLAB code. It will check for any defects or cracks or any missing tablet from the strip with the help of the image that was taken through camera. Based on the quality check performed using image enhancement and by carrying out morphological operations in MATLAB, it would signal the microcontroller to start the conveyor belt if no defect is found and in situations of faulty tablet motor 2 is signal to run a rejecting mechanism. A LCD is connected to get the count of number of strips checked, rejected or for the ones that are found defective. USART is used for communicating with the PC through microcontroller and vice-versa. 1.10 In the end, I could complete the project on time. The most difficult part of my project was to design and propose the circuit diagram. As working with PCB designing software was completely new for me. It took several trials before coming on the final figure. And implementation of the hardware was tough. Few minor mistakes created led into a big problem in my end mistake. I de-soldered the wrongly connected component part and connected it in the proper way to get the desired output and the circuit running. At the last stage, a test run was performed before submitted to check the working of each component and to see that the expected output is achieved. Working in teams was a challenging task as coming on one mutually agreed statement and solution is quite difficult. Though I managed my team well and never made any unfair decisions. 1.11 SUMMARY Thus, for companies manufacturing number of tablets in a day, inspecting the final product manually can be tiresome and a tedious job. The company need the help of automated systems. So, my project is an economical and efficient way to solve this problem. They only problem with this system is that it cannot be used for transparent tablet strips and is made to use for single colored tablets. This problem can be solved by performing further expansion on my project. By performing this project, I learnt a way in which the image processing technique is applied. It was a great learning experience. I could use my engineering knowledge and use it to apply theoretical knowledge into practical grounds. While conducting surveys, I could understand the market position and market value of new technologies. It gave me a good exposure. [S1]