Not all fog removers are equal: Unmasking the impact of dehazing on object detection

zDehazing is an important branch of computational photography aiming to enhancing image clarity by removing atmospheric haze and scattering effects, crucial for improving visibility in applications such as unmanned aerial vehicles, traffic control, and autonomous driving.However, most of the studies in this particular field lack an assessment of the developed algorithm in context of object detection (OD).In this study, we aim to quantify and evaluate the contribution of several stateof-the-art dehazing methods (C2PNet, D4, Dehamer, gUNet) on OD using YOLOv8, known for its superior performance.For this purpose, we utilized the test portion of the VisDrone-DET dataset including 548 haze-free aerial images as the data source.For a more comprehensive assessment, we evaluated these approaches to object detection under different haze levels and resolutions.Since it is inherently impossible to obtain hazy and clean images simultaneously, we (1) generated synthetically hazed images involving varying haze densities and (2) resized to 640p and 1280p resolutions.Next, we used YOLO8 and YOLO10 models to evaluate the OD performance in (i) haze-free ground truth, (ii) three different hazed versions, and (iii) their dehazed counterparts through several metrics.Our experiments showed that the gUNET approach, incorporating a variant of the U-Net model inspired by GCANet and GridDehazeNet outperformed the others in terms of OD performance.Surprisingly, the Dehamer negatively affected the OD performance due to the artifacts it produced.This assessment not only provides valuable findings into the effectiveness of these methods but also sheds light on how to benefit them when it comes to object detection under hazy atmospheric conditions. Sis giderimi insansz hava aralar, trafik kontrol ve otonom sr gibi uygulamalarda hayati nemdeki grnrl iyiletirmek amacyla atmosferik pus ve salm etkilerini ortadan kaldrmay hedefleyen hesaplamal fotografinin nemli bir daldr. Ancak bu alandaki almalarn birou gelitirilen algoritmann nesne tespiti (NT) balamnda deerlendirilmesinden yoksundur. Bu almada stn performansyla bilinenYOLOv8 zerinden son teknoloji rn eitli sis giderici yntemlerin (C2PNet, D4, Dehamer, gUNet) katksnn NT balamnda llmesi ve deerlendirilmesini amalanmtr.Bu amala veri kayna olarak VisDrone-DET veri kmesindeki 548 sissiz gkyz grnts ieren test ksmndan faydalandk.Daha kapsaml bir deerlendirme iin farkl sis seviyeleri ve znrlkler altnda NT balamnda bu yaklamlar deerlendirdik.Sisli ve temiz imgeleri doal olarak ayn anda elde etmek mmkn olmadndavn, (1) deien sis younluklar ieren sentetik sisli imgeler oluturduk ve (2) 640p ve 1280p znrlklerinde yeniden boyutlandrdk.Ardndan (i) sissiz kesin referans, (ii) farkl sislendirilmi srm ve (iii) bunlarn sisi giderilmi muadillerinde YOLO8 ve YOLO10 modelini kullanarak NT performansn eitli ltler zerinden deerlendirdik.Deneylerimiz GCANet ile GridDehazeNet'ten esinlenen