Sistem za delno avtomatsko štetje polipov na slikah
| Autor: | Bohte, Boštjan |
|---|---|
| Přispěvatelé: | Kononenko, Igor |
| Jazyk: | slovinština |
| Rok vydání: | 2015 |
| Předmět: |
computer and information science
računalništvo model visokošolski strokovni študij accuracy slika computer science večnivojski perceptron pretreatment error računalništvo in informatika diploma označevanje napaka diplomske naloge udc:004.85(043.2) multilayer perceptron točnost polipi image predobdelava polyps labeling |
| Popis: | Cilj diplomskega dela je implementirati sistem za delno avtomatsko štetje polipov na fotografijah morskega dna. Dobili smo veliko fotografij, pri čemer ročno označeni smo vzorec teh uporabili za učno množico, na podlagi te pa smo izvedli strojno učenje. Fotografije je bilo treba predobdelati z ustreznimi metodami, ki so popravile napačno osvetlitev, in tako bolj poenotiti vse slike. Najprej smo naučili model, ki odstrani ozadje na sliki, pri tem pa pusti vse, kar je del polipa. Sliko z odstranjenim ozadjem ponovno predobdelamo za model avtomatskega označevanja polipov. Uporabnik se nato odloči, ali bo avtomatsko označene polipe popravil in s tem dobil točno število polipov, ali pa bo to prepustil modelu napovedovanja števila polipov. Ta je naučen tako, da predvideva napako prejšnjih modelov in ustrezno napove bolj točno število polipov. Celotni sistem smo testirali nad testnimi slikami in ugotovili, da nekatere slike niso primerne za celotni sistem ter zato jih ne moremo uporabljati v tem sistemu. Pri ročnem popravljanju avtomatskih oznak polipov smo ugotovili, da je občutno hitrejše od ročnega označevanja polipov na celotni sliki. Model napovedi polipov je bil tudi pozitivno ocenjen, saj v povprečju zmanjša napako modela avtomatskega označevanja polipov. The objective of the thesis is to implement a system for semi-automatic counting of polyps on the photographs of sea-bed. Numerous photographs have been obtained, wherein the manually labelled sample of those was used for the train set, and on basis of this the machine learning was implemented. Photographs had to be re-processed with appropriate methods in order to correct the bad lighting and furthermore even out the differences. First a model was used, which removes the background in the image while leaving everything that is part of the polyp. Image with the removed background is re-processed again for the model for automatic labelling of polyps. The user can then decide whether automatically labelled polyps will be repaired and thereby the exact number of polyps is obtained, or leave it to the model for predicting the number of polyps. This model is taught to foresee the error of previous models and to predict the number of polyps more accurately. The entire system was tested by test images and it was established that some images are not suitable for the entire system, and, therefore cannot be used in this system. While manually correcting automatic labels of polyps it was established that it is significantly faster than the manual labelling of polyps in the overall photographs. The model for predicting polyps was also positively assessed, as it reduces the error of the model for automatic labelling of polyps on average. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|