computer vision v.3

8/7/2019 Computer Vision v.3

http://slidepdf.com/reader/full/computer-vision-v3 1/24

Bedy Purnama, S.Si., M.T.



5W + 1Hy Wh aty Wh y y Wh erey Wh eny Who

y How



Cl ip_Terminat o r



Tinjauan Computer Visiony C o mputer Imagingy Akusisi & pemr o sesan visua l Inf o rmasi co mp

y C o mputer Visi o ny

Aplikasi image digunakan

comp

y Image Pr o cessingy Aplikasi imaga digunakan manusiay I. Resto rati o n, I. En h ancement, I. C o mpressi o n

y Image Ana l ysisy Melibatkan pemeriksaan data citra untuk

memfasi litasi peme cah an masa lah pen citraany I. Segmentati o n, I. transf o rms,

feature extra ctio n & pattern classificatio n



D efinisiy Goal : membuat keputusan yang bermanfaat tentang o bjek fisik

nyata dan lingkungannya berdasarkan apa yg di tangkap citray C omputer Vision men co ba meniru H uman Visiony C omputer Vision = pro ses o to matis yang mengintegrasikan

sejum lah besar pr o ses untuk persepsi visua l, seperti :y Akuisisi citray Pengo lah an citray Klasifikasiy Pengena lan ( R ecognition )y Membuat Keputusan.

y Vision = Geometry + Measurement + Interpretation



P roses dalan Computer Visiony Pro ses-pr o ses da lam computer vision dapat dibagi

menjadi 3 aktivitas :

1. Memper o leh atau meng akuisisi citra digita l2 . Melakukan teknik k o mputasi untuk mempr o ses atau

mem o difikasi data citra ( o perasi- o perasi peng o lah an citra )3. Mengana lisis dan meng interpretasi citra dan menggunakan

h asil pemr o sesan untuk tujuan tertentu, misa lnyameng o ntr o l pera latan, memantau pr o ses manufaktur,memandu r o bo t, d ll .



P roses dalan Computer Vision

(cont d)



RISET MILITER INDUSTRI



RISE TComputer Vision For The Blind ( Indoor Navigation Through Traje ctory Tracking )



INDUS TRIy C ust o m_So ft w are_t o _Inspe ct_Measure_ w ith _

Video

y Identify_ Cl assify_and_Ana l yze_Images_in_Vide o _and_P ho to s

y So ft w are_Deve lo pment_Servi ces_f o r_Digita l_Image_Pr o cessing

y So ft w are_Pr o gramming_Servi ces_f o r_Image_Reco gniti o n_Feature



M ILITER



G rafika C itra

Pengo lah an C itra Lanjut

C o mputer Visi o n



Kajian dari Computer Visiony Image ana l ysis met ho ds

y edge dete ctio n, feature dete ctio n, image segmentati o n,

Image transf o

rmatio

nsy Mo tio n and vide o ana l ysisy Metho ds f o r reco nstru cting t h ree-dimensi o na l scene

inf o rmati o ny C o mbining rea l imagery w ith co mputer generated

imagery y Virtua l Reality, Augmented Rea lity



Image analysis metho ds (cont d)y image segmentati o n

y Image transf o rmati o ns



M otion an d video analysisy G esture Dete ctio n

y C alibrati o ny Tracking



M etho ds for re constru cting 3D

scene information



Combining real imagery with

computer generate d imageryy Virtua l Reality

3D sensed data can be

studied f o r surgi cal pat h s to be f o ll ow ed by a surge o n o r aro bo t.

In th e future, rea l-timesensing and registrati o n canbe used f o r feedba ck in t h epro cess.



Combining real imagery with computer generate d imagery (cont d)

y Augmented rea lity : vie w s o f rea l o bjects + augmentati o n

Cl ip 01, 02 , 03

perso n wo rks at rea l deskremo te co ll abo rato r

represented by pi cture o r video o r talking h ead

o bjects o f discussio n; e.g. a

patient s brain image, migh

talso be fused int o visual fieldHOW IS TH IS A C H IEVED?

Fro m University o f W ash ingt o n H IT Lab



y KK IC My Sub KK Multimediay Bedy Purnama & Tj o ko rda Agung B W

computer vision v.3

Documents