Measurement error1794
Pages:
1
|
bonduell private msg quote post Address this user | |
| Hi there! Here is trouble occurred with camera sensors. I have an order 3D-shoot. It is restaurant on TV tower in Vilnius, Lithuania. I have no idea how I can fix this bugs. Tower is a circle. Take a look please. Is there anybody here who can do some advice? Thks. |
||
| Post 1 • IP flag post | ||
|
Max_Sodomovskiy private msg quote post Address this user | |
| Привет. Покажи скриншот Capture. Может быть, причина в зеркалах на потолке. Google Translate Hey. Show Capture a screenshot . Perhaps , the reason the mirror on the ceiling. |
||
| Post 2 • IP flag post | ||
|
|
bonduell private msg quote post Address this user | |
| Не похоже, высота уходит с какой то конкретной точки и постепенно, это хорошо видно на obj модели если ее поставить ребром. А вот почему стены не сошлись я вообще не понимаю. Google Translate Swap to RussianFull screen close It does not look like the height of the leaves to a specific point and gradually , it is clearly seen in the obj model if it is to put an edge . And that is why the wall did not agree , I do not understand . |
||
| Post 3 • IP flag post | ||
|
JohnBecker private msg quote post Address this user | |
| Bonduell, I don't see a problem. What exactly is not working properly? | ||
| Post 4 • IP flag post | ||
|
|
Max_Sodomovskiy private msg quote post Address this user | |
| То есть при съемке проблем с выравниванием не было? Да, стены и стол продублировались. Зеркальный потолок. Думаю, что надо повторно загрузить. Google Translate That is, when shooting problems with alignment was not? Yes , walls and desks duplicated . Mirrored ceiling. I think that it is necessary to re-download . |
||
| Post 5 • IP flag post | ||
|
|
bonduell private msg quote post Address this user | |
John, if you look to 3D model you can see that start point and finish point inside circle is different about 2 meters. In real world it's not true ) The circle is not connecting to self  |
||
| Post 6 • IP flag post | ||
|
|
bonduell private msg quote post Address this user | |
| Max, в capture режиме тоже видно что стены не сошлись. Проблема именно в самой съемке. может быть допуски датчиков сыграли странную шутку. я недоумеваю. написал в службу поддержки, но не думаю что они что то дельное ответят Google Translate Max, in the capture mode is also seen that the wall did not agree . The problem is in most shooting . Sensor tolerances can be played a strange joke . I am perplexed . I wrote in support , but do not think that they have something sensible answer |
||
| Post 7 • IP flag post | ||
|
|
Max_Sodomovskiy private msg quote post Address this user | |
| Ты снимал по часовой стрелке, если смотреть сверху и закончил в том месте, где сейчас ошибка? Есть ли возможность переснять этот участок, двигаясь в обратном направлении. Для МП это хороший вопрос, явный баг съемки. Потолок. Google Translate You shot in a clockwise direction when viewed from above , and ended at the point where the error now? Is it possible to re-shoot this area , moving in the opposite direction . For MP is a good question , a clear bug shooting . Ceiling. |
||
| Post 8 • IP flag post | ||
|
|
bonduell private msg quote post Address this user | |
| Да, так и было. снимал по часовой и пришел на 2 метра выше. ну надо понимать что съемка производилась на высоте 200 метров от земли Google Translate Yes, it was. I filmed in a clockwise and came to 2 meters higher . well, we must understand that the shooting was carried out at an altitude of 200 meters above the ground |
||
| Post 9 • IP flag post | ||
|
|
Max_Sodomovskiy private msg quote post Address this user | |
| Хорошо, что там нет датчиков высоты Надо идти в обратном направлении. Google Translate Well , there is no height sensors ) must go in the opposite direction . |
||
| Post 10 • IP flag post | ||
|
|
bonduell private msg quote post Address this user | |
| Там есть гироскоп, и это его проделки Google Translate There is a gyroscope, and this is his antics |
||
| Post 11 • IP flag post | ||
 WGAN ForumFounder & WGAN-TV Podcast Host Atlanta, Georgia |
DanSmigrod private msg quote post Address this user | |
| English or English + Russian Please | ||
| Post 12 • IP flag post | ||
|
|
bonduell private msg quote post Address this user | |
| Dan, ok. Could you do advice about this trouble? | ||
| Post 13 • IP flag post | ||
 StandardMember Wandsworth U.K. |
RobinLycka private msg quote post Address this user | |
| @bonduell - Assuming you moved around starting from one point and doing a full 360, maybe it would work starting from one point and adding one scan to each side making the splice at 180 degrees rather than the full 360. That should just mathematically give it less of an inaccuracy. Did you try Matterport support to see if they can fix it ? (Mind you, judging by the comments I´ve seen, I don´t think they´ll be to quick to respond). |
||
| Post 14 • IP flag post | ||
|
360Verbeelding private msg quote post Address this user | |
| Hi, i think the problems occure because of the windows and table setting. There is almost no differnts. So the MP cam sees almost every where the same image. Its like scanning a long hall, everything looks the same. There are also a lot of "extra" scans that you don't need for the walk tru, hide those. There is probably just one or two scans that cause the problem. You can find those by starting with the first scan, hide this scan, press publisch and see what happens. If nothing changes, unhide and hide the next scan, publisch again and look again. This way you might find the problem. If you found the faulty scan, delete it and you schould be fine. Happy scanning, Grtz Rene. |
||
| Post 15 • IP flag post | ||
|
WGAN Forum Founder & WGAN-TV Podcast Host Atlanta, Georgia |
DanSmigrod private msg quote post Address this user | |
| @bonduell Windows may be acting as mirrors, if so: 1. Duplicate model 2. Use mirror markers for Windows in affected area 3. Reprocess 4. Let us know if this fixes 5. Let us know what Matterport says If the camera was in direct sunlight in the affected area: 1. Duplicate model 2. Delete-not hide-scan in the affected area (always helpful to overscan a space so you have this option) 3. Re-process model If ceiling or floors acting as mirrors: 1. Duplicate model 2. Delete scans in affected area 3. Reshoot affected space with different lighting (different time of day) when rescanning, you will need overlapping scans be fire you get to the affected area 4. Reprocess It is an amazing space and perfect for a Matterport Spaces 3D tour. Makes us want to visit you  Dan |
||
| Post 16 • IP flag post | ||
|
|
bonduell private msg quote post Address this user | |
| @Dan Thank you guys. Thank you Dan for your reply personally ) I will be waiting for MP response. I don't want to do second shooting, it takes more than 4 hours. You can see more info about this tower and about Howner project here: https://howner.com/v/show/247KB4 https://howner.com It is platform for matterport pro and agents. Still beta, but no matter. It would be nice if you do feedback about project Howner. There is realtime chat inside between agents and guests. And many more ideas inside my head Looking forward to your reply! |
||
| Post 17 • IP flag post | ||
|
CarlosFHdz private msg quote post Address this user | |
| I think that alignment issue is related to a similar conversation with the capture app misalignment allowance increasing. That is a huge difference overall. I would recommend when doing something similar to go back and forth from one side to the other to build the circle from a middle point. That way there may be a better alignment at the end. Instead of heading in one direction for the whole circle. |
||
| Post 18 • IP flag post | ||
|
|
bonduell private msg quote post Address this user | |
| Good news guys! MP has fixed my model! They reprocessed it manually. They said nothing about trouble and reasons. Thks! |
||
| Post 19 • IP flag post | ||
Pages:
1This topic is archived. Start new topic?