|
|
Automatic acquisition of values from measurement devices without communication interface
Dohnálek, Martin ; Čala, Martin (referee) ; Kunz, Jan (advisor)
This bachelor thesis deals with the matter of optical character recognition from displays of measurement devices without communication interface. This would allow carrying out automated experiments using cheaper or older gear, which is not endowed with means for direct connection to a computer. Input image necessary for the character recognition is acquired using a camera pointed at a display of the device. The recognition is afterwards performed on periodically captured image based on an already existing dataset for particular apparatus. The output of the algorithm is a file containing recognized values, units, and timestamps of the recognition. The tool for creating datasets was designed as well. The achieved speed of recognition (as fast as 34 ms per iteration) during practical testing confirmed the sufficient optimalization of OCR algorithm. On the other hand, the determined hit rate of recognition abiding specified conditions was nearly 100 %. Lastly, the resistance to misalignment of display and sensor plane was monitored. The OCR algorithm is resilient to horizontal tilt up to +/- 5° and vertical tilt up to +/- 20°.
|
|
|
Optical granulometry
Dohnálek, Martin ; Richter, Miloslav (referee) ; Kunz, Jan (advisor)
This diploma thesis deals with the topic of macroscopic particle size detection during freefall using computer vision. The measurement results can be used to verify other granulometric methods, or in the industry as a simpler alternative to conventional methods. Particle sampling takes place in flight and particles may overlap. Because of this, the data is recorded by a high-speed camera on the scene with fit-to-purpose lighting. The components of the proposed algorithm are: background removal, localization of potential particles, segmentation using the watershed method, determination of particle size. Using image correlation, the fall velocity and the number of particles in the sample are also determined. Experiments confirmed that the measured particle sizes of the sample correspond to the sizes determined by the traditional sieving method and are consistent even on repeated measurements. The correctness of the speed determination was verified by a theoretical calculation. When determining the effect of the sample’s dustiness on the quality of detection, resistance to the presence of dust up to 10% of the sample weight was found.
|
|
|
Automatic Acquisition Of Values From Measurement Devices Without Communication Interface
Dohnálek, Martin
This paper deals with optical character recognition (OCR) of measured values from displaysof measuring instruments without communication interface. Proposed algorithm functions asa bridging between a screen of the instrument and a measuring software. It enables the acquisitionof the displayed value from a snapped picture from the input camera stream, so that it is comprehensiblefor computers. The execution is, after necessary initialization done by the user, fully automated.Supported camera connection interfaces are USB and WiFi, meaning that either standardoffice webcam, or smartphone with third party app running is usable. The final algorithm recognizedboth 99 % valid values and its speed was exceeding the refresh rate of most common instruments(as fast as 34 ms per iteration). This means that it is not bottlenecking the measurement itself.
|
|
|
Automatic acquisition of values from measurement devices without communication interface
Dohnálek, Martin ; Čala, Martin (referee) ; Kunz, Jan (advisor)
This bachelor thesis deals with the matter of optical character recognition from displays of measurement devices without communication interface. This would allow carrying out automated experiments using cheaper or older gear, which is not endowed with means for direct connection to a computer. Input image necessary for the character recognition is acquired using a camera pointed at a display of the device. The recognition is afterwards performed on periodically captured image based on an already existing dataset for particular apparatus. The output of the algorithm is a file containing recognized values, units, and timestamps of the recognition. The tool for creating datasets was designed as well. The achieved speed of recognition (as fast as 34 ms per iteration) during practical testing confirmed the sufficient optimalization of OCR algorithm. On the other hand, the determined hit rate of recognition abiding specified conditions was nearly 100 %. Lastly, the resistance to misalignment of display and sensor plane was monitored. The OCR algorithm is resilient to horizontal tilt up to +/- 5° and vertical tilt up to +/- 20°.
|
guest ::
RSS feed.