Observability of the relative motion from inertial data in kinematic chains

Kok, Manon; Eckhoff, Karsten; Weygers, Ive; Seel, Thomas

doi:10.1016/j.conengprac.2022.105206

Cited by 3 publications

(1 citation statement)

References 39 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…2D tasvirga va raqamni nuqta kuzatishga asoslangan algoritmlardan foydalanadigan yechimlar kinematik zanjirning aniq tasviri (qirralar bilan bog'langan nuqtalar to'plami) natijasida yuzaga keladigan xatoliklarning to'planishi bilan bog'liq muammolarga ega. Shunday qilib, Manon Cock, Carsten Eckhoff, Yves Wagers, Tomas Seale [13] tomonidan taqdim etilgan ishda oyoqqa biriktirilgan giroskopik sensorlar bilan tajriba misolida skelet modelini namoyish etadi. Tahlil skelet modelining ikkita qattiq bog'langan tugunlari yordamida amalga oshiriladi.…”

Section: Kirishunclassified

Skelet modeli asosida kompyuter ko‘rish tizimlarida inson harakatini tekshirish

Beknazarova,

Jaumitbaeva

2024

Международный журнал теоретических и прикладных вопросов цифров

View full text Add to dashboard Cite

Kompyuter ko‘rish tizimlarida inson harakatini tahlil qilish ikki o‘lchovli va uch o‘lchovli fazolarda amalga oshiriladi. Ikki o‘lchovli fazoda qiyofalarni tahlil qilishning o‘ziga xos xususiyati bitta kameradan yoki bir nechta tana sensorlaridan olingan tasvirlarni qo‘llashdan iborat. Bunday yondashuv xatoning tez to‘planishiga olib keladi va natijada rasmning aniqligi past bo‘ladi. Uch o‘lchovli tahlilni amalga oshirishda bir nechta kameralar qo‘llaniladi va qiyofalar hajmli elementlar to‘plami sifatida taqdim etiladi. Uch o‘lchovli tahlilning yuqori aniqligiga qaramay, u yuqori hisoblash murakkabligi va ichki kompyuter tarmog‘idagi yuk bilan bog‘liq. Tadqiqot maqsadi bitta kamera tasviridan foydalangan holda modelni ishlab chiqishdir. Ko‘rib chiqilayotgan modelda inson qiyofasi sifatida tasvirlangan skelet, tasvirlangan assegmatik grafga asoslab hisoblash. Inson qiyofalarning tahlil tuzilmalari bajarildi va o‘n besh asosiy ball ta’kidlangan. Jismoniy va mantiqiy aloqalarning matematik tavsifi berilgan va tadqiqot qilingan. Nuqtalar va ulanishlarning tezligi va fazoviy xarakteristikalari harakatning umumiy dinamikasini tavsiflaydi. Inson harakati modelining tavsifi va uning muayyan tasvirlar uchun qurish modeli keltirilgan. Taklif etilayotgan model asosida grafik tugunlari va qirralarining nisbiy joylashuvi va tezlik xarakteristikalari haqidagi ma’lumotlarni to‘plash va tahlil qilish algoritmi ishlab chiqilgan. Ishlab chiqilgan model inson harakatining etalon (normal) dinamikasi haqida ma’lumot to‘plash uchun ishlatilishi mumkin. Agar standartdan chetga chiqadigan dinamika aniqlansa, xatti-harakatlar deviant sifatida aniqlanadi. Taklif etilayotgan algoritm inson harakatini aniqlash va tahlil qilish, shuningdek, harakatlarning mos yozuvlardan og‘ishlarini aniqlash uchun kompyuter ko‘rish tizimlarida qo‘llanilishi mumkin.

show abstract

Section: Kirishunclassified

Skelet modeli asosida kompyuter ko‘rish tizimlarida inson harakatini tekshirish

Beknazarova,

Jaumitbaeva

2024

Международный журнал теоретических и прикладных вопросов цифров

View full text Add to dashboard Cite

show abstract

慣性センサを用いた水平加速度の積分に基づく歩幅推定

NAKANISHI,

SONOBE,

HIROSE

2023

Transactions of the JSME (In Japanese)

View full text Add to dashboard Cite

The variation in stride length is an important index of gait stability. The evaluation requires a device that can accurately measure the stride length during ground walking. However, there are currently no practical and reliable measurement systems for clinical use. The purpose of this study was to improve the stride estimation method using a single inertial sensor and to verify the accuracy of stride length estimation. Stride length was estimated by integrating the horizontal acceleration at the ankle joint. To reduce the impact acceleration at ground contact, which is the primary factor causing estimation errors, we mounted an inertial sensor on the lower leg. In the present method, the following four factors contribute to the stride length estimation error: the sensitivity of the inertial sensor, consideration of the directional angle, detection of integral intervals, and estimation of the relative position vector from the inertial sensor to the ankle joint. This study presents solutions to these problems and verifies the accuracy of stride length estimation by comparing it with optical motion capture measurements. The results indicated that the sensitivity of the inertial sensor improved the accuracy of the stride length estimation. Furthermore, we found that the standard deviation of stride estimation per step was approximately 32 mm, regardless of cadence and stride length. We conclude that the stride length estimation by the acceleration integral method is applicable to the estimation of the average stride length; however, it is difficult to apply it to the evaluation of stride variation from step to step.

show abstract