Ukubona into yinto ebalulekileyo yeComputer Vision, apho injongo kukuchonga nokubeka izinto emfanekisweni. Enye yeendlela zokubonisa indawo yento emfanekisweni yiBhokisi yokuBonding. Ibhokisi yokubopha yibhokisi yoxande enokuthi ibalwe ngendlela elula ebandakanya unciphiso lwemathematika olusisiseko kunye nomsebenzi wokwandisa.
Ibhokisi, ngaphezu koko, inokumelwa zilungelelaniso ezimbini, i (x, y) yekona ephezulu ekhohlo kunye ne (x, y) yekona esezantsi ekunene. Olu lwazi lungqina ukuba lubalulekile kwizicelo ezahlukeneyo zobomi bokwenyani, ukunceda iingcali ukusuka kwabo bacuphayo ukuya kushishino lweemoto eziziqhubayo.
Inkcazo yeNgxaki kunye nesisombululo
Ingxaki ephambili esijongene nayo kumfanekiso kunye nokufunyanwa kwezinto yindlela yokuchonga indawo yento emfanekisweni ngokuchanekileyo. Isisombululo kukusebenzisa ibhokisi yokubopha, enokubalwa kusetyenziswa indlela elula ebandakanya iilayibrari ezahlukeneyo zepython.
Python lolona khetho lubalaseleyo kulo msebenzi kuba unamathala eencwadi atyebileyo kunye nezixhobo ezenza lula inkqubo, ziyenze isebenze kwaye iqonde. Amathala eencwadi amabini aphambili asetyenziswa rhoqo-OpenCV kunye neMatplotlib.
Indlela ye-OpenCV kunye neMatplotlib
I-OpenCV imele ithala leencwadi loMbono weKhompyutha uMthombo oVulekileyo kwaye ibandakanya amakhulukhulu ee-algorithms zombono wekhompyuter. IMatplotlib, kwelinye icala, yilayibrari yokucwangcisa yolwimi lwenkqubo yePython kunye nolwandiso lwezibalo lwamanani iNumPy. Inika zombini indlela ekhawulezayo yokujonga idatha evela kwiPython kunye namanani omgangatho wokupapashwa kwiifomathi ezininzi.
import cv2
import matplotlib.pyplot as plt
# read image
image = cv2.imread('input.jpg')
# our bounding box coordinates
box = (x1, y1, x2, y2)
# Draw rectangle (bounding box)
cv2.rectangle(image, (box[0], box[1]), (box[2], box[3]), (0, 255, 0), 2)
# Display the image with bounding box
plt.imshow(image)
plt.show()
Umfanekiso ulayishwa kusetyenziswa indlela ye-imread esuka kwi-cv2, kwaye ibhokisi yokuda izotywa kusetyenziswa cv2.uxande umsebenzi othatha umfanekiso kunye nezilungelelanisi ezimbini ezimelwe 'yibhokisi'. Iiparamitha ezimbini zokugqibela zinemibala kunye nobukhulu ngokulandelelanayo. Le khowudi iya kubonisa izinto ezikumfanekiso wakho zibotshelelwe ngokupheleleyo yibhokisi.
Ukusetyenziswa kweeBhokisi zokuBonding
Ukuqukumbela, iibhokisi ezibophelelayo zidlala indima ebalulekileyo kwimisebenzi yombono wekhompyuter, kubandakanya ukufumanisa into, umbono wekhompyuter, kunye nokulungiswa komfanekiso. Banikezela ngesisombululo esisebenzayo nesisebenzayo sokufumana izinto kunye nolwazi lwemethadatha ngaphakathi kwemifanekiso. Ukufunda ukuphumeza ngokuchanekileyo iibhokisi zokubopha kwiPython kunokuba luncedo kakhulu kuye nabani na obandakanyekayo kuphuhliso lwesoftware, ukufunda ngomatshini, okanye i-AI yobulumko bomsebenzi. Ayiloncedo kuphela kukhuseleko kunye nokucupha, kodwa ikwanceda kakhulu kwizicelo ezinjengokubona ubuso kunye nokuqondwa, ukubhaqwa kwabahambi ngeenyawo, kunye neenkqubo zoncedo lwabaqhubi abaphambili (ADAS) kwiimoto eziziqhubayo.
